While there is quite a lot of information on EFI conversions for our cars (some of which I will directly reference here), my goal with this article is to help anyone embarking on this type of project with a modular approach, so that one may go at his own pace, and deviate for personal preferences at any point along the way. My own project has been done on a 1975 base 2002 using Megasquirt 2, Ford EDIS, and B&G firmware, so this will be the basis referenced here. All standard disclaimers apply, please be safe about working on your car, and I'm not responsible if you screw something up, but I hope this helps many people interested in pursuing various EFI conversions for their 2002!
Useful 2002 Megasquirt conversion blogs:
Intro - Some things to consider
If you're just contemplating this undertaking and wondering if it's a good project for you and your car, here's my brief personal advice on the subject. But obviously do your research and decide if it's a project you think you'd like doing.
Reasons EFI is a good fit for your 2002:
- You like tinkering with your car on a regular basis
- You like to drive your car regularly, and thus value better driveabiltiy and reliability
- You have and are irked by issues with chokes, cold starts, and warmup
Reasons EFI is NOT a good fit for your 2002:
- You value originality (your car might end up as much E30 as 2002 by the end).
- Your main goal is performance (there are easier, cheaper, and quicker paths to pure horsepower).
- You prefer to have someone else work on your car (this can make the tuning process slow, cumbersome, and frustrating).
- You want it done quickly or are worried about scope creep. (You WILL find other side projects you'll want to do along the way; the project WILL grow and take more time and money as it goes along. This isn't necessarily a bad thing, but you need to be prepared for all of the 'well, while I'm in here' offshoots).
First things first. . .
You may not yet have decided on all of the details for a project with such a large scope yet, but that's actually OK, because there are some basic upgrades that make sense to take care of first, and are basically necessary no matter what direction you end up going with your car. We're going to start with a few upgrades to the coolant and electrical system. These items can, and I would say even should, be made to 2002s regardless of EFI, but are certainly required for EFI, and therefore make a good starting point.
Section I - Coolant System Preparation
We'll begin with the cooling system, for which two specific upgrades are needed:
First is an E30 coolant divider with three sensor ports. Try to pick up a used one here on the FAQ, Ebay, or a junkyard, and ideally get one that has all of the E30 sensors already in it. You'll replace one with the stock 2002 sender to keep your dash gauge working, use the stock E30 sensor to provide coolant temperature data to Megasquirt, and leave the third in as a plug or use it for other needs (such as a switch to drive an electric cooling fan). The additional coolant sensor is critical for fuel injection, as it is the main input that adjust how much fuel is injected during a cold start and then during warmup until the car is up to operating temperature. Conveniently, it can also be used as the input to have Megasquirt drive an electric radiator fan directly, which is really nice. Here's what this coolant divider looks like in my car, with the 2002 sender up front, the temperature sensor for Megasquirt behind it, and the switch/plug on the left:
Next up, while the coolant system is open, will be the coolant bypass line and hose, also from an E30. You'll need this for pretty much anything other than the stock coolant manifold, be it side draft carbs, ITBs, or the 318i intake. You can salvage one from an E30, or buy a brand new one (or as part of a complete set with all new coolant hoses) from Ireland engineering here: http://www.iemotorsport.com/bmw/2002-cooling/M10sddrfthrdwr.html
And here's what it looks like installed on my car, sans intake manifold:
On last thing that is worth mentioning here. While you have the coolant drained for performing these upgrades, it makes this a convenient time to also take the radiator out. I'll leave scope-creep items like radiator and fan upgrades for other articles/blogs, but the reason I mention it here is that IF you are planning to use a Ford EDIS ignition setup (which is my recommendation, but see the later ignition section for more details), you will need the radiator out so that you can remove the crank pulley and replace it with one with a trigger wheel. The easiest route that I would recommend is to just purchase a brand new pulley with a trigger wheel and also the sensor mount from Tom at 02again (http://www.02again.com/?page_id=358).
Section II - Electrical System Preparation
So, with the main items for the coolant system in process and/or already taken care of, we will next move to the primary electrical system upgrades needed for EFI. The first involves relocating the battery from the original location in the engine bay to *somewhere* else in the car. The most popular new homes are either in the trunk or under the rear seat, but you can put it pretty much wherever you want to, as long as you get it the heck out of the way up front. I didn't like the idea of losing trunk space and drilling into the rust-prone rear shock towers, so I chose to follow Zeebucks lead and installed two Hawker Odyssey batteries under the back seat, and will link to his complete instructions for this here:
The only deviations I made from his method were to route the cable through the interior and through the drivers side firewall instead of underneath the car, and I then brought the positive terminal into a sealed junction box on the inside of the front drivers side fender just underneath the relay bracket. Here's a great picture I nabbed while I happened to have the engine out:
I'm pretty sure I picked this up at a local Lowes or Home Depot, but I haven't been able to find it again. So at least here's a link to something similar that I did find on Amazon:
I further followed in Zeebuck's footsteps and installed the larger 80-amp alternator from a 318i, in order to have enough overhead to power all of the additional electronic components for EFI and engine management systems I'd be adding. Again, his guide for this is already complete and excellent, so I'll point you to that write up here:
Lastly, I added a small additional blade fuse box (picked it up either from Autozone or Amazon: https://www.amazon.com/OLS-6-Way-Blade-Indicator-Protection/dp/B00QMTAZ1W) up in the front of the car to provide the fused terminals for the forthcoming additional electronics. The +12V supply for this fuse box is provided from a relay which is switched by the ignition. The stock 2002 ignition switch actually powers a LOT of things directly and needs to handle fairly high current. This isn't exactly desirable, and you certainly don't want to any more load to this poor 40 year old switch, so please heed this advice and use a relay for this job, and do NOT power any additional electronics directly from the ignition switch. On square-light vehicles, you'll be looking for a solid green wire from the ignition switch to run the coil (terminal 85 or 86) on the relay. You can find this wire several places; it supplies fuses #4 and #12, powers the stock ignition coil, and runs to the lights and turn signals.
Something else I should add here is that the stock wiring for the headlights is not that great, as the headlight switch must switch the full load for the headlights, some 20 amps! Although there are relays in the circuit, they aren't used as relays should be. As such, I took the opportunity to rewire the entire relay area and put in a new relay box to house everything. I mention this not because it's necessary for EFI, but because it's what you'll see in all of my pictures and wiring diagrams, so you need to know what you're looking at. Here's my full wiring schematic and final product, for those interested:
One final note in this section, and that's regarding electrical connections. You can continue to use plain old spade terminals for just about everything, but as this project involves a lot more wires and connections than the original system, I found it easier to buy a kit full of Weatherpack connectors to make multi-wired connections. My new fuse box for example uses two 5-wire connectors (white in the photo) to hook up the 10 wires coming from the relays to the wiring harness. (I think I ran the ground wire individually). If you want to take a similar route, here's a good Weatherpack starter kit on Amazon:
Section III - Wideband O2 sensor
This is the final stand-alone part that is key to running EFI, but also equally useful for tuning a carburetor-equipped car. There are two basic types of O2 sensors, narrow- and wide-band. Narrow band sensors check if there is any un-burnt oxygen in the exhaust stream and report that back as either a rich or lean condition. Wideband sensors on the other hand are a little more sophisticated and report back just how much rich or lean the engine is running. The target here is an air/fuel ratio in the range of 12~15 (depending on exact conditions) which represents the point at which both all fuel and all oxygen are burned. For either a carb or EFI, this feedback lets you see under which operating conditions you should change the amount of fuel flow to your engine to try to maintain this perfect balance all the time. With EFI, it's as easy as adjusting the numbers in the tuning software, and for a carb it means trying out some different jets. It can also be used in EFI for closed-loop feedback, where the engine management system will automatically make fueling adjustments on the fly based on what the oxygen sensor is seeing.
I installed the fairly common LC-1 wideband kit from Innovate:
There are many others, some probably better and some worse, but I chose this one as it was not hideously expensive and because it provides two analog outputs, one of which goes to Megasquirt and the other which can drive the gauge in the cockpit. The important points to remember for installing an Oxygen sensor are:
- Mount the O2 sensor in the upper half of the exhaust pipe, at a point AFTER the exhaust streams from all 4 cylinders have come together. Here's what mine looks like right by the front of the transmission:
- Make sure to program BOTH of the two analog outputs for 0-5V. As received one of the two outputs is programmed to function like a narrow band instead of a wideband.
- I mounted the control unit on the passenger side of the engine bay just in front of the firewall. You can see it in this picture zip tied to the lip just above the distributor area:
Section IV - Ignition
(NOTE: If you are only interested in ignition control and want to keep your car carbureted, there is a system called Megajolt which is similar to Megasquirt but only for ignition control without EFI:
Ok, now we have reached the first major decision point in the project. It's time to decide what type of ignition system you'd like to run. I'm going to assume that you want SOME type of electronically controlled ignition, as this project would pretty much be a waste otherwise. Here are the three main options that I know of to choose from:
1.) 123/TUNE - https://www.123ignitionshop.com/gb/tune-bmw/106-123tune4rvbmw.html
This option has the advantage of being stand-alone, you could install this straight away on a stock car and be done if you wanted and not bother with EFI. It also very stock looking, and works wonderfully with a Bosch blue coil. If you looking for something quick and turnkey, this is what I would probably recommend. However, if you plan to continue on with a Megasquirt EFI installation, I would instead recommend going with one of the next next two options.
1.) Megasquirt direct coil control - This is basically the same as the 123Tune setup, with the only differences being you will instead use your stock distributor (it's only purpose here is to direct the spark to the correct plug), and you will program the timing through Megasquirt, which will then directly control the firing of the ignition coil.
NOTE: you will need to make sure your Megasquirt control board is built properly to support this. Here's the link to the direct coil control section in the MegaManual:
3.) Ford EDIS - Don't let the Ford brand name dissuade you, this is a truly elegant and awesome ignition system! It is a distributor-less system that works in a wasted-spark configuration, so the only inputs it needs are crank position and RPM, and it can be easily driven by Megasquirt with whatever advance curve you desire. If you're going EFI, and especially if you're using Megasquirt, this is the way to go. The parts are readily available, usually quite cheap, and it integrates perfectly in with MS. I will detail out its parts and installation below.
NOTE: you will need to make sure your Megasquirt control board is built properly to support this. Here's the link to the EDIS section of the MegaManual:
4.) Coil on Plug (COP) - You may by all means pursue and COP system and drive it with Megasquirt, and I think several 2002 owners have done so. But my advice on this one is frankly don't bother and just stick with the Ford EDIS. I won't dissuade you if COP is what you want, but I feel there's basically no advantages for using it in a 2002, and here are my reasons. COP is mainstream today because it offers several advantages on MODERN cars. These include things like no HT wires (which interfere with AM radio reception), longer dwell times to make a more powerful spark, and the ability to control the ignition on each individual cylinder, including things like multiple ignition events. But in order to control just one cylinder at a time, we need to know which one of the two paired cylinders (2 or 3 and 1 or 4) is on it's compression stroke vs. exhaust stroke. This generally requires a camshaft position sensor (as far as I know impossible to install on a 2002), or requires the signal from the distributor (to identify which cylinder should be firing), but to me that's just hokey to still have the distributor as part of a distributor-less ignition setup. The way around this, and how the Ford EDIS works, is to utilize what is known as a wasted-spark configuration, where the spark is fired for both paired cylinders (based just on crankshaft position), and the spark for the cylinder on its exhaust stroke is therefore 'wasted.' COP can be set up in a wasted spark configuration also, but now the advantages of longer dwell time and individual cylinder control are negated, and it's really no different than the Ford EDIS setup, just more complicated. Unless of course, you strongly value AM radio.
Here's a link to the best thread I've come across on this topic, should you decide you'd like to research the topic further:
Installing MS-controlled EDIS in the 2002
If you'll take my recommendation, this is the way to go, and here's how you accomplish this:
1.) Read and understand the EDIS section in the MegaManual:
2.) Trigger wheel - As I mentioned in the coolant section, the easiest way to do this is to buy the correct crank pulley with the trigger wheel already on it from 02Again (http://www.02again.com/?page_id=358). Sadly, this option was not available to me when I started my project, so instead I had a local shop turn a collar for me to mount a scavenged Ford trigger wheel to a stock E30 crank pulley. But I can tell you the next time that I have the radiator out of the car for some reason, I'm going seize the opportunity to upgrade to the 02again pulley/wheel! Anyway, here's how mine looks like currently:
If you look closely in the picture, you can see a small white arrow on the wheel where one tooth is missing, and this is lined up with the #1 cylinder TDC mark on the crankshaft. This missing tooth tells the EDIS brain where TDC is, and then it 'counts' each tooth as it moves past the sensor so that it always knows what position the engine is in, and can decide when to fire the spark accordingly. The sensor
should be mounted so that is lined up with the 5th tooth AHEAD of the gap when the missing tooth is at TDC. There's a 'trigger offset' parameter in the Megasquirt software to adjust this if it's not perfect, but it's wise to still try to get it pretty close, so that the EDIS will function correctly in 'limp home' mode. This is a built-in backup where, if for some reason there is no signal from Megasquirt as to what timing is desired, the EDIS will default to simply firing consistently at a static 10deg BTDC. This means the car will still be able to run on just EDIS alone, albeit not very well at higher revs and at a loss of power, but it's great for testing to make sure everything is working and just in case something should go wrong.
NOTE: If you have or desire air conditioning in your car, that complicates matters as the compressor pulley is right were we want to mount the trigger wheel. I think this is still possible, but will likely require a different trigger wheel and some extra ingenuity and custom work on your part.
3.) Crank position sensor - Should be pretty obvious by this point, this is the VR sensor that senses the teeth on the crank trigger wheel and sends the signal back to the EDIS brain. There's a link to the right sensor on the above mentioned 02again website, and the right connector I know can be sourced here:
The only two things to remember here are to make sure that the sensor wires are shielded to prevent electrical noise in the signal, and to set gap between the sensor and the wheel teeth to about 1mm.
4.) EDIS module, coil pack, and wires - Honestly your best bet for the module these days is probably Ebay, although if you have a local salvage yard that you like to frequent, look for an early 90s Ford Escort/Mercury Tracer to liberate these parts from. You'll want the EDIS-4 module, connector, and coil pack connector (don't bother with the donor coil pack or plug wires themselves, see below). Should look like this:
The original Ford coil pack and wires are ludicrously difficult to mount, but fortunately more user-friendly brand new options are available for cheap, such as this coil pack from Amazon for just $20:
To install this coil pack, I *think* I purchased this wonderful mount from 02Again, but I don't see it listed on the website, so you might need to inquire. It mounts in the stock distributor location, nicely plugging the hole for the now unnecessary distributor while maintaining a stock-ish look:
For plug wires, the *RIGHT* set to look for is a 2001-2003 Ford Taurus 3.0L V6 with 24V/DOHC. This will fit both that coil and the 2002 cylinder head. Do NOT get wires from the very similar 3.0L SOHC V6 from the same vintage Tauruses!!! Here's what I bought:
Here's how my wires look installed, note the use of the E30 exhaust manifold gasket with the extra exhaust heat shield:
5.) Wiring - This is pretty straight forward, just follow the diagram below. EDIS pins 1 and 3 to Megasquirt pins 24 and 36, VR sensor goes to EDIS pins 5 and 6, both paired shields go to pin 7, pin 8 gets +12V (from the new fuse panel), 9 is ground, and 10 and 12 go to the coils:
NOTE: This is important to keep your tachometer working! The Megamanual presents a schematic using some diodes to tie the output of the two coils together to drive the tachometer, but I worked for months on this and could never get it to work properly. I think the flyback voltage that the stock tach wants to see is higher than what makes it through the diodes. What DID work for me in the end is actually much simpler. EDIS pin #11 (CTO) is the tach signal output. This with a big NON-POLARIZED capacitor (I used 0.068uF) in the line directly drives my tachometer perfectly throughout the whole rev range! In this picture, you can see where I mounted the EDIS module to the firewall and you can even see the orange capacitor dangling down just below the module on the yellow wire, before it plugs into the original factory tachometer wiring:
Section V - Megasquirt controller
This will be a pretty short section, as you only have two major decisions to make here: Which version of Megasquirt and do you want to buy: a turnkey pre-assembled module or the kit and build/solder your own? I built my own; mostly for the fun of it, but it was also a bit cheaper. But if you don't like soldering or are in a hurry, it's probably worth the extra $200 to buy the pre-assembled version. There may be other sources, but the main one that I know of and would recommend for all Megasquirt kits is www.DIYautotune.com. Here's a short list of the options and my thoughts on each of them:
Megasquirt I - This is the cheapest option at only around $200 for the kit, and it does in fact have all the capability necessary to run a naturally aspirated 2002 withEFI and spark via EDIS, making this a perfectly acceptable route for a budget build. I would however in general recommend stepping up to MS2 for most people, primarily because the MS2 community is larger and therefore it's easier to find answers than for MS1. I also think the MS1 processor is now obsolete and no longer supported.
MS1 DIY kit: https://www.diyautotune.com/product/megasquirt-i-programmable-efi-system-pcb3-0-kit-w-black-case/
MS1 assembled: https://www.diyautotune.com/product/megasquirt-i-programmable-efi-system-pcb3-0-assembled-unit/
Megasquirt II - This was my pick because it's far cheaper than MS3, has all the capability you could ever need for a 2002 (including forced induction, etc), and a nice large support community. As far as assembly time, I think it took me about a month working about an hour or so at a time several evenings a week. It's definitely time consuming, but an absolutely tremendous learning experience for both the principals of EFI/engine control as well as electronics in general, which is what made it worth it for me. I feel that the knowledge gained here makes the tuning process vastly easier to tackle when that time comes.
MS2 DIY kit: https://www.diyautotune.com/product/megasquirt-ii-programmable-efi-system-pcb3-0-kit-w-black-case/
MS2 assembled: https://www.diyautotune.com/product/megasquirt-ii-ems-system-smd-pcb3-57-assembled-ecu
Microsquirt - This is basically the same thing as a pre-assembled MS2 but in a smaller package and slightly cheaper. I think the one drawback is that it needs and additional module to support idle control with a stepper motor. Since I've already had the fun and learning of building one MS2 setup, I would give Microsquirt some serious consideration if I were ever to do a second car.
Megasquirt III - While the capabilities of MS3 are truly awesome, I think it's really hard to justify the additional cost for use on an M10 engine. Features like 8-cylinder sequential injector control, 4-bank wankel control, water injection, nitrous, CAN-bus support, etc. are just, well, unnecessary for a 2002. But if for some reason you are interested in going this route, here's a link.
MS3 assembled: https://www.diyautotune.com/product/megasquirt-iii-ems-system-v3-57-assembled-unit-w-black-case/
The last main thing you'll need is the MS wiring harness, which I STRONGLY recommend you just buy instead of make. For $80, you get the right connector complete with 10 feet of high quality, different colored and labeled wires. You just can't beat that!
Once you've sourced or built your Megasquirt board, don't forget that before sealing it up in it's enclosure that you'll need to load some firmware on it. There are two main types of firmware:
1.) 'Stock' firmware (use MegaTune software for tuning)
2.) MS/Extra firmware (use TunerStudio software for tuning)
I used the stock code and MegaTune and now that I'm used to it and have the car running great I'm not going to bother changing, but for new builds I would recommend using the MS/Extra code and TunerStudio. It recent years it seems to have 'won out' in the mainstream and for all intents and purposes, it's just better.
Instructions and source for stock firmware: http://www.megamanual.com/ms2/install.htm
Instructions and source for MS/Extra firmware:
Optionally, you might find it useful to also pick up the Stimulator. This neat little doodad runs on a 9V battery and plugs into the Megasquirt controller and simulates all of the various engine systems (e.g. RPM, MAP, temperature, AFR, etc.) This allows you to fully bench-test and program your Megasquirt so that you know it basically works BEFORE you start hacking into your actual car. Below is a picture of my just-completed MS2 on it's first test run using the Stimulator. Boy, I can still remember bouncing off the walls with happiness that evening!
Section VI - EFI Hardware
Here's where things start to get fun, installing the major components needed for EFI, but this is also sort of a point of no return, so make sure you have your Megasquirt controller working, all of the parts ready, and enough down time lined up before you pull the trigger. We'll start with the list of parts/hardware needed, and then go into the details of each:
1.) Intake manifolds
2.) Throttle bodies
3.) Individual throttle bodies (ITBs)
4.) Fuel rail & injectors
5.) Fuel pump
6.) Additional sensors
7.) Idle control
8.) Megasquirt controller and wiring harness
NOTE: Plan out and source everything before installing anything, and then start with the wiring (step 8.) first!
1a.) Plenum intake manifolds - For a more tii look, I think it's possible to use an E21 320i intake manifold or even a 2002tii manifold, but I'm not going to recommend that as I have no idea how to get the right fuel rail or injectors for it, but I think it has been done before. For 95% of us, the E30 318i intake is the way to go. It's basically plug-n-play, and is fantastically engineered by those Bavarians for off-the-chart fantastic mid-range performance. I'd guess less than $50 on ebay or from junkyards. Just one personal request, please take the time to clean up and paint your manifold. It's easy to do before installing it, and makes it everything look so much better!
2.) Throttle body - If you pick up a manifold with the 318i throttle body already on it, then great! For a stock to mildly-modified engine this will work just fine. Megamanual calculator says this should be fine up to 116 horsepower. If you plan for your engine to go above that however, you'll probably want a slightly larger one from either the 325e or 325is. Here's a link to a detailed run down of each and, as always, there's a wonderful adapter plate available at 02again.com!
While you're shopping on 02again there are some other accessories that you'll probably want to order as well, including the IAC adapter (for idle control), the throttle position sensor mount, and again though this isn't on the website I think I sourced this nice set of aluminum plugs from there for plugging up all of the various unused ports on the TB. One note, I did need to backfill some of the openings with epoxy, as the plugged holes would whistle something fierce at certain throttle positions!
3.) - ITBs - For those that desire more top end horsepower than mid-rage torque, there are several ITB options for EFI out there. Having previously loved dual DCOEs, I've often toyed with this idea, but for me I think it will need to be done on a different car. There simply isn't enough room in the 2002 engine bay to get long enough runners on ITBs to match the mid-range performance of the impeccably designed 318 manifold, so in my opinion ITBs are a better match for an engine build that's designed for a >4000 RPM power band. But if you have high compression pistons (10.0:1 or greater), a rather aggressive cam (292 or greater), and some porting, this will likely be the route you want to pursue. I know of two vendors (formerly TWM, now Borla, and Dbilas) that market EFI throttle bodies with mounting geometry that matches DCOE carburetors, so if you already happen to have a manifold for dual sidedrafts, this becomes rather straight forward.
The TWM/Borla parts are found here: http://www.borlainduction.com/2900-series.html and are probably the way to go if you already have a sidedraft manifold.
If you don't already have a manifold, then I would probably go for the Dbilas kit, as it comes with everything including the throttle bodies: http://www.dbilas-shop.com/Products/Throttle-body-kit/Street/BMW/M10/Mutli-throttle-intake-system-for-1602-1802-2002-316-318-518-520-E21-E30-1-5-2-0-8V-M10::10351.html
Lastly, I have also toyed with the idea of running just one of these throttle bodies on the Lynx single-sidedraft manifold that I have. I've done some calculations and think it should work, but would take some pretty specific selection of injectors and fuel control setup. If you want to experiment with this, contact me directly for the specifics as I don't want to bore everybody with the math here, but in case this peaks your interest, here's the link to the manifold:
I won't go into much more detail on ITBs here as there is already a good writeup on this from Johnup, so see here for further reading on the subject:
4a.) Fuel Rails - Unfortunately I can't offer much help with fuel rails for E21 or tii manifold setups, but the others are easy; the ITB vendors all supply their own rails and the 318 intake uses the stock 318 rail including fuel pressure regulator. Again you can clean up and use the one from a donor vehicle or buy new parts. Here are links to the right parts from ECS, but I think local dealer pricing is also decent on these parts, with the added bonus of offering a CCA discount:
Fuel rail: https://www.ecstuning.com/b-genuine-bmw-parts/fuel-rail/13531707731/
Fuel pressure regulator: https://www.ecstuning.com/b-bosch-parts/30-bar-fuel-pressure-regulator/13531722040~bos/
Injector retainer clips: https://www.ecstuning.com/b-genuine-bmw-parts/fuel-injector-clip-priced-each/13531274729/
(DON'T FORGET THESE CLIPS, otherwise the rail can pop off of the injectors, dumping high pressure fuel into you engine bay, ask me how I know!)
4b.) Injectors - You'll need to estimate your engines peak horsepower for this, but once you do that the rest is easy. The important thing to remember is to get the SMALLEST injector that you can which still flows enough fuel at peak horsepower. The reason not to oversize much on the injectors is because then the pulse width will get very very short at idle, making it both difficult to tune and not as smooth of an idle as is possible with the smaller sized injectors. Here are my guidelines:
Up to 120HP get 19# Bosch yellow tops: https://www.fiveomotorsport.com/bosch-yellow-top-19lb-fuel-injector
120-140HP get 21# Bosch pink tops: https://www.fiveomotorsport.com/bmw-0280150440-pink-top-13641703819
140-160HP get 24# Bosch blue tops: https://www.fiveomotorsport.com/24lb-bosch-fuel-injector-0280150947-blue-top
Here's the correct EV1 connector for all of the above Bosch injectors: https://www.diyautotune.com/product/fuel-injector-pigtails-bosch-ev1/
And lastly here's the fuel injector in the MegaManual for more information:
5.) Fuel pump - We need to pause here and again give thanks to the old generation of Bavarian engineers, who developed stuff that just plain worked and then didn't feel the need to mess with things every few months just for the hell of it! It is because of this that we have the very great fortune of having a high pressure fuel injection pump available that drops straight into the 2002 fuel tank with no modifications whatsoever! Obviously the E30 is once again our benefactor, and ebay and junkyards are the best budget options. The pump is available new and not too expensive from off-brand manufactures, but the sending unit for the fuel gauge (which ALSO works perfectly with the 2002 gauge, as hard as that is to believe) I only see listed as 'genuine BMW' for lots o' $$$. Here's the link to ECS with the various options:
Note: there are two versions of the hanger, one with a return fitting and one without, so check to see if your tank has a return fitting on the tank itself. If it does, here's the version with only the supply fitting:
You will need to install another relay to power the fuel pump (power for it can come from that nice new fuse panel), and this relay's negative coil will be controlled by Megasquirt. I did something pretty clever with the wiring here: since my rear window defrost wires were all rotted and non-functioning, I used those existing defrost wires to run back to the fuel pump instead of trying to run a new set of wires back through the whole length of the car. All installed, mine looks like this:
Other than a short priming pulse at start-up, MS will not run the pump unless there is an RPM signal >0. While this is moderately safe, additional oil pressure and/or roll-over sensors can be added as further safety measures to cut off the fuel pump in the event of an accident.
And as always, here's the link the fueling section of the Megamanual:
IMPORTANT: You MUST run new fuel line rated for high pressure fuel injection for the supply line from the tank to the fuel rail! If you have a late-model Sqaurie, you can cheat a little by swapping things and using the steel RETURN line on the driver's side of the car for the supply, and then running the low pressure return through the plastic line the runs through the passenger side interior. Be sure to run all new fuel injection rated rubber lines everywhere on the high pressure supply though, and certainly do NOT use the stock plastic line for the supply!
6.) Sensors - EFI requires a few extra sensors than what were normally included on cars in the 70s, so here's the list of additional input sensors that you'll need to plan on adding:
a.) Coolant temperature - See previous coolant section, use the E30 temp sensor in the coolant divider neck, and program the temperature curve in Megasquirt.
b.) Throttle position sensor - Use the TPS sensor from a late 80s Nissan 300Z with the adapter mount from 02again.com
c.) Oxygen sensor - See above, use and Innovate LC-1 or similar heated wideband O2 sensor with 0-5v analog output.
d.) Inlet air temperature sensor - Use this open element GM sensor somewhere in the intake track before the throttle body and program temperature curve in Megasquirt:
7.) Idle control - The two basic options for idle control are a fast idle solenoid and a stepper motor idle air control valve. If you've got the 318i intake route, your job is again ridiculously easy as you can just get the adapter block from 02again.com (I told you at the beginning this was a great place for this project!) and the proper 90s Jeep stepper motor. Wire it up according the the Megasquirt wiring diagram for a stepper IAC and here are the 02again installation instructions:
If you are not using the 318i manifold, then a fast idle solenoid is likely easier to install as it can be plumbed in with hoses, but this pretty much exhausts my knowledge of fast idle solenoids here, and I also unfortunately have no idea how to approach idle control with ITBs.
On a side note, if you use the stepper motor, Megasquirt allows you to configure your unused fast idle output for other purposes; I have mine set up to control my electric cooling fan!
Megamanual link to idle control:
8.) Megasquirt controller and wiring harness - Maybe I should have put this further up, because these are really the FIRST pieces of hardware that you'll want to install, on the other hand you need to already have planned out all of the above hardware and options before installing anything. Anyway, once you know what hardware you're going to use, how everything will be plumbed, where all of the sensors will go, and so on; you should then start laying out the wiring harness. It's easiest to do this with all of the old hardware out of the way so you can really work at tucking the wires back out of the way to keep things clean looking. I recommend laying out all of the wires individually, and then wrapping them/sheathing them in conduit in sections. Obviously if you haven't already ridded your car of all of the obscene 70's emissions junk, then do this concurrently, as reuse those same wiring clips for the new stuff. The Megasquirt controller needs to be mounted in the car's interior (it's not meant to survive in the engine bay environment), so this means making one big hole somewhere in the fire wall big enough to pull the whole wiring harness through. The best spot I found for this was on the passenger's side high up in the foot well, just underneath where the brake lines come across (obviously be careful not to drill through your brake lines!) I also used a small bit of coolant hose as a grommet:
NOTE: Since you'll likely have purchased a wiring harness with the Megasquirt DB39 connector already on one end, this means you'll need to feed the whole length of the harness through the firewall from inside to outside, and this takes some patience!
Besides the harness, you also need to run a vacuum line through the firewall to get Megasquirt the MAP signal. I was able to do this right along side the plastic fuel return line, thereby saving drilling an additional hole.
Once the whole harness is through the firewall, you can start routing all of the individual wires. Once you have everything laid out, the next step is to install all of the proper connectors on the ends, and then finally wrap and tuck all of the various branches. Here's how my finished wiring harness looked laid out in the engine bay before installing any of the hardware:
It's trickier than it sounds to layout wiring cleanly, so take your time with this process, make good soldered connections, cover them in heat-shrink tubing, and wrap/tape up conduit joints and ends, and you'll end up very pleased with how much cleaner the whole engine bay looks in the end! Here's a copy of the general Megasquirt wiring diagram (including EDIS) that's a good one to work from:
As I mentioned the MS controller box itself needs to be mounted in the interior of the car. Many people choose to mount it inside the glove box which works just great, but I didn't want to give up that much real estate in my glove box so instead I chose to mount mine on top of the transmission tunnel behind the center console where, as long as the car has no A/C, affords a good bit of space. I did later end up also installing an amp for my built-in Android tablet here and an RS232-to-Bluetooth adapter for wireless connectivity to Megasquirt (https://www.efianalytics.com/products/class1Bluetooth.html), at which point I moved the MS module up on to the face of the heater core box (having already had the heater box out once, that's not a job I plan to EVER do again on this car!)
Original Authors: Trent Tilton and Rob Shisler
Transmission seals kit
Shifter rebuild kit
New six- or eight-bolt guibo with new nuts and bolts
Exhaust flex-sealing ring
Clutch kit (including e21 323 throwout bearing)
New copper exhaust downpipe nuts (3 of them)
Slave cylinder extension hose
New set of flywheel/pressure plate bolts
We are working on getting part numbers for all these parts.
Probably one of the most "transforming" bits of work you can do to an '02 is to remove the standard Getrag 232 four-speed transmission and convert the car to accept a G-245 five speed overdrive transmission from the later '80-83 e21 320i.
The overall operation is not that difficult, although it may seem daunting at first. It is a task that can be done by someone with basic mechanical skills in their driveway (although it is preferable to do it on a level surface). For this article, Trent used a kit he got from Dave Varco at www.2002parts.com. Dave is a great guy, and the kit was complete with all the parts listed above without the need to do any fabrication. There are other suppliers who provide kits, or you can make your own kit, but Trent chose Varco's kit because it is complete, priced right, and he happens to be a friend of Trent's. Rob, on the other hand used a DIY setup he pulled from an already-converted parts car, adding some of his own modifications (more on this later).
In terms of sourcing a good candidate five speed tranny, the best bet is to just head off to a local boneyard. There are no guarantees with a junkyard tranny, but when compared to the cost of a rebuilt one (at least $600-800 in parts and tools to rebuild yourself or at least twice to 3x that for a pro-built box.) Reputable '02 shops will sell you a known good box for around $350, but Trent chose to roll the dice on a boneyard tranny. He found one at a local yard on a "Half Price Day" for only $75. Even if it is junk, and he has to try again, I'm still saving a little bread for the trade-off of uncertainty and hassle of swapping in a second five-speed, but only if needed.
Check to make sure that the seals on your boneyard box are not "ovalized" as this is transferred into the metal housings. This will make it impossible to change the seals with out the new ones leaking. When you get the "new" tranny home, clean it thoroughly with your favorite solvent or degreaser, and order a complete seals kit for it from your favorite parts vendor.
You will also want to get a shifter rebuild kit ordered, as well as a new guibo and center bearing. In addition, you will need a new exhaust flex-sealing ring, clutch kit, e21 323 throwout bearing, new copper downpipe nuts, gasket, and possibly a new speedo cable, slave cylinder, and slave cylinder extension hose. [unfortunately, we do not have these part numbers right now, but a call to a knowledgeable parts house should give you the right parts. -Rob]
It might also be a good idea to use new guibo nuts and bolts, as these can be subjected to some flexing and metal fatigue over the years. A new set of large flywheel bolts (the ones that hold the flywheel to the crank) is also recommended, but make sure you get the right kind for the size of your flywheel (either 215mm or 228mm).
Drain the tranny with a 17mm hex tool or with one you made out of an old VW or similar lug bolt with a nut welded on the other end or a pair of nuts jammed together.
You will definitely want to change all the seals on the tranny itself. The fact is they tend to leak after so many years. This includes the rear main seal which requires a thin wall 30mm socket (use an impact wrench to air it off -- or have a shop do it) and change out the seal with a seal puller (or just have the shop do it).
Be very careful with the new seals. Put a thin layer of RTV on the outside and some moly grease on the inside edge for starter lubrication. When refilling it, there are a few options. Some prefer the standard GL-4 (non-hypoid) gear oil while others prefer synthetic gear oil like Redline MTL.
There is some controversy over synthetics, but rest assured MTL is a good choice, especially if you've gone to the trouble of using new seals. Even ATF can be used with good results. Be aware that some have reported new leaks with synthetics because they can actually clean out gunk that might have been stopping or slowing down leaking spots. Either way, fill the tranny to the point where it begins to come out of the fill hole, then put the plug back in. (Do this with the car or transmission leveled.
This is also a good time to rebuild the shifter linkage. All the plastic washers and joints should be replaced. (Check the FAQ Index for this one.) Remember to put back in the exact number in the exact location. This will give your shifter a nice tight feeling. These parts are shared on most BMWs and a good independent shop or dealer should stock them.
In terms of the conversion itself, the first thing you need to do is SAFELY elevate the car and put it on jackstands. Ideally you would pick a flat hard surface in a level location (such as inside your garage). If this is unavailable you can do it in the driveway, but make sure the e-brake is on (and working) and/or you have securely chocked up the rear wheels. Start by putting a 2x4 across the cradle of your floor jack to spread the weight a little and lift the car at the center of the front subframe. If you are on a soft surface (NOT RECOMMENDED!), use thick plywood or equivalent to keep the jack and stands from digging into the surface. Place the a couple more 2x4s on top of the stands and then position them under the stamped steel frame rails attached to the front floors of the car. Then, if you are on a flat surface, you can raise the rear of the car in similar fashion, placing the stands under each end of the rear subframe, just next to the mounts. Make sure the whole thing is rock-steady before crawling underneath! Use good quality stands and/or ramps because cheap ones can collapse! Remember: Safety First!
Once the car is safely in the air, the first thing to attack is the exhaust system. There are three bolts at the bottom of the exhaust downpipe that will allow you to move the exhaust out of the way. When you remove the down pipe from the manifold, you will need a new gasket. Trent didn't need to totally remove his because it is a custom job that runs under his rear subframe, but on stock cars you will need the room. If your downpipe nuts are rusted on try soaking them with Liquid Wrench and/or blasting them with a torch. If they are REALLY bad, you might need to cut the nuts off (ouch!) with a Dremel or similar tool. Then remove the four Guibo nuts and bolts. Remember to note their direction for reinstallation. Then remove the four bolts that bolt the driveshaft to the differential. Next, remove the two 13mm bolts that hold the center bearing to the tunnel. Watch your head because at this time the driveshaft might fall on it! Next, remove the bolts that hold the transmission to the engine block. Don't forget the stamped metal shield at the bottom of the bell housing, otherwise the tranny wont slide off the flywheel/clutch assembly. Support the rear of the transmission with a jackstand, and then remove the bolts holding the tranny crossmember to the tunnel. Having an assistant at this point is very handy. Next, unhook the linkage at the bottom of the shifter. Finally, carefully slide the tranny off and let it thud to the ground. (Maybe put a stack of magazines under it just to avoid chipping the garage floor-the tranny itself is resaleable if the synchros are still good. Or there is always fleaBay!)
Next, remove the old clutch/pressure plate by removing the bolts holding them to the flywheel. You may also want to have the flywheel resurfaced at a machine shop, while you're in there replacing things. Make sure the shop knows to put a .020" "step" on the perimeter of the flywheel so that the pressure plate holds the clutch tight and centered with the intended amount of pressure. While the flywheel is being done, test fit the new tranny to the engine to determine how much and exactly where you need to clearance the tunnel for the slave cylinder. In most cases, a couple good whacks with a big hammer (try to use a non-marring "dead-blow" version) will move the sheet metal enough to make room for the new slave cylinder. Remember that motors move around during spirited driving, so a little "elbow room" is needed. Some kits (like the one from Varco) recommend that you put the slave on after the tranny is already in the car, but the opposite actually appears to be the easier way to do it. Make sure the bleeder valve is pointing down when you install the new slave!
At this point we recommend that you make or buy a tranny cradle for your floor jack. It will make your life a hell of a lot easier during the next stage of the conversion. One way is to take a piece of 2x6 wood and then screw or nail a couple of pieces of angle-cut 2x2 on each edge to create a kind of V shape that will steady the tranny enough for you to jack it into position. It is also very helpful to have a friend man the jack while you guide the tranny itself into position.
The five speed tranny is exactly 3.6" longer than the four speed it is replacing, which means we need to move the crossmember mounting tabs back in the tunnel exactly that amount. You will also need to shorten the driveshaft by this amount, or buy a kit with a shortened driveshaft in it. The Varco kit I used came with a new set of bolt-on tabs for moving the crossmember and also the shortened driveshaft. You can also cut the mounting tabs out of a donor car (e21 or 02 - it doesn't matter), then grind the tunnel metal off them and then either weld or bolt them to the tunnel in the correct spot. Remember that the elevation is as important as how far back you place the new tabs. To place the tabs with bolts, bolt the transmission to the motor with a couple of bolts and then attach the driveshaft by just snugging up a couple of those bolts too. Put a level on the bottom of the driveshaft on each side of the center bearing to make sure the angles are the same and the transmission is not positioned too high or too low. Take your time at this point because it is critical Drill the holes and bolt the brackets up.
Another way to solve the move-the-mounting-tabs problem is to do what Rob did and make a set of rails from some 1" angle iron from a hardware store (or old bedframe) and cut them to fit between the original four-speed mounting tabs and the center bearing mounting tabs further back in the tunnel. Then the cross member just attaches to the rails in the correct spot, while the elevation is just about spot-on perfect just by attaching the cross member directly to the bottom of the rails. If you need to, you can adjust the rear of the tranny down by using some big washers in between the crossmember and the rails, or even up with a little by stacking a (big) washer or two on top of the rubber tranny mount itself.
In Trent's case, drilling the holes for the new mounting tabs required pulling back the carpet inside and obviously drilling new holes in the tunnel. The rail method eliminates this alteration of the car's sheetmetal but requires more fabrication of the rails themselves from blank angle iron, as well as some grinding for clearancing the speedo cable. You may also need a little more hammering on the tunnel where the rails run from original crossmember mounting points and the center bearing mounts. Finally, the rails will not hit the original tabs "square" due to the change in elevation from original tranny mounts to center bearing mounts, but this is curable by either cutting the "L" of the angle iron to allow it to bend to fit, or just by tightening the bolts down really hard to force the mounting tabs to bend a little to meet the rails squarely. Its your call on what to do here, but since part of the point of the rail method is to avoid altering the car too much, Rob recommends bending the rails themselves.
Once everything is fitted correctly, disassemble everything and remount the flywheel and clutch. The flywheel bolts must be changed with new ones and then torqued to 75 lb.ft. Use a small amount of red locktite on the threads just for luck. Don't put too much because it affects the torque number you need to use due to its lubricating effect during installation. If your new clutch kit didn't come with a centering tool, you can cut the old input shaft off the four speed box and use that. Center the clutch, then carefully bring the new tranny into position. Spin the output flange while engaging the clutch splines to make it easier to engage them. Once all lined up, the tranny should just slide home and fit tight against the block. If it wont slide all the way home, pull it off and check for something in the way or not aligned correctly and try again.
After the transmission is bolted up, it is time to just finish up all the little things. The new speedo cable, wire up the reverse lights, bolt the driveshaft back up, re-install the shifter, put the carpet back (if needed), then take for test drive!
If you have any questions don't hesitate to post them to the Forum. Trent would like to thank his dad for his help, and Dave @ www.2002parts.com for the kit and great advice!
Written by John Aho Monday, 05 September 2005
Installing Exhaust Manifold or Header Studs
By John Aho
Studs are designed to be fitted only "finger-tight" into their threaded hole in a casting. However, a locking type of nut (PTN, or prevailing torque nut) is often specified to go on the stud. These two facts present an inherent dilemma. The grip of the PTN on the stud can easily turn the stud when you don't want it to (how often have you removed an exhaust manifold nut and had the stud come out at the same time?)
With this in mind, the installer may be tempted to over tighten the stud into the hole, to make certain it stays put. This mistake happens a couple of ways: Either the stud gets torqued after it bottoms in the hole, or the stud's shoulder (unthreaded section in between the threaded ends) jams into the surface at the edge of the hole. Some studs have no shoulder (all thread), and some holes have no bottom (through hole). This allows for another problem: with nothing to stop it, the stud can insert too deep.
So, why is it wrong to tighten the stud into the hole? Unfortunately, I have even seen it done by supposed "professionals". The way a stud can outperform a bolt in clamping efficiency has to do with the even distribution of stress across the engaged threads. If you torque a stud into a blind hole, you lose that benefit, and concentrate stresses at the first thread in the hole. If you torque the shoulder of a stud against the edge of a threaded hole, you lose the benefit & usually distort the mating surface. It's not as if there is a suitable bearing surface to take the load. Just don't do it.
Loctite threadlocker is the answer for securing the stud (for most of our situations), but it's worthwhile to look at the design intent of the purely mechanical fit. Studs require a "fit classification" that provides a very close fit. Common American thread classes include #2 (free fit, general application) and #3 (very close fit). Classes #4 & 5 are actually interference fits, and you won't find them readily available.
Metric thread fits are classed differently, using a number to designate "tolerance grade", and a letter to indicate "fundamental deviation" (upper case letter if internal thread). When a class of fit must be provided for mating threaded parts, the tolerance symbols of the internal thread is shown first, then a forward slash, followed by the external thread's symbols (for example, 5H/4h). Basically, ISO metric "6" approximates a UNF class #2 fit. The metric range also includes 4 & 5 (closer) and 7 & 8 (looser). The letter H or h indicates deviation allowance = none.
You can get a decent "ball park" on fit class by the feel of threads engaging, but to really measure it is much harder. The major diameter (O.D.) of the stud's thread is not the primary determining factor for class of fit. The pitch diameter (an imaginary cylinder located where the thread width is equal to the space between adjacent threads), which requires a special "thread micrometer" or the 3-wire system to measure, determines the class of fit. The bottom line is you should try to only use metric steel studs designated as 4h. Good luck finding them.
The '02 parts book shows M8 x 40 studs, but the length used can vary from 38 to 45 mm depending on availability and/or thickness of the header flange. Be sure to select studs with a shoulder. Customarily, metric steel studs have their nominal length (which protrudes) less than their overall length by approximately 1.25 x diameter. This would mean about 10 mm of thread goes into the head. The last BMW studs I got had about 15 mm to thread in the head, which is even better. The blind holes on the bottom row of the exhaust side are about 20 mm deep, so you won't bottom out as long as you put in the correct (short) end. There's some decent manifold studs used by other manufacturers, if you look around. The Saturn cars use the same size, and have a feature on the outer end that is made to fit a female E-7 torx socket.
So what does all this mean to me when I've got an exhaust stud backing itself out of a buggered hole on a 2002 head? You will have to clean out the hole completely to get the Loctite compound to work properly. Getting this done on the car, with the manifold still in place, is a pain. Even with the exhaust out of the way, the top row of stud holes go right through into the head, making them very hard to get clean & dry.
To proceed: the loose studs will come right out, and you can "double-nut" to remove the rest.
Take off the valve cover and blot away the oil around where the top four threaded holes go through.
Carefully use a proper thread chaser to clear out any old threadlocker or other crud.
Clear out the debris you've just shoved into the head before it washes down into the sump.
Use an aerosol can of brake cleaner to wash out the threads, use a rag pushed tight up against the inside of each through hole to prevent solvent/debris from getting inside. Please Note: The Loctite compound will only work if you do a good job getting things clean & dry. Follow the instructions that come with the product. Try to find Loctite 262 (a bearing supply house will stock it), otherwise use the readily available 271. Again, use as directed: just one drop on each clean, dry, new stud as they go in.
After the Locktite has set, reinstall the exhaust with new gaskets, and torque the 8 new copper plated steel PTN to 30 Nm (22 ft lb).
Original Author: Jeff Ireland
Differential Types - 2002 and 320i:
Early 1600's and 2002's came with a long neck differential. These were phased out in early 1969. They were either 4.11:1 (1600) or 3.64:1 (2002). They are very rugged (heavy) and easy to come by, but not rebuildable due to the lack of available replacement parts. If you want to switch to the later short-neck differential, you must replace the entire subframe and rear suspension. This sounds tough but is actually fairly easy and the parts are readily available.
All later 2002's and all 320I's used the short neck differential. These came in several ratios:
3.36:1 -- 2002 Turbo
3.45:1 -- some euro tiis and 6 cyl E21s
3.64:1 -- most 2002s and all 4-speed e21s
3.90:1 -- some 1976 2002s and all 5-speed e21s
4.11:1 -- all 1600s
Of all these variations, all of them can be transplanted into a 2002.
Variations and how to recognize them:
All 2002/1600 diffs have 4 bolts holding the side covers and the output flanges are held in by a large central bolt. CV's are attached the output flanges with 6 - 8mm bolts.
Early e21 ('77 and '78) 320i's have four bolts holding the side covers, but the output flanges are held in by a C-clip inside the diff. These diffs usually have 8mm bolts for the CV joints but some use the later 10mm bolts. These early diffs are the same width as the 2002 diff so no spacers are necessary if you are putting one in a 2002.
Later e21's have 6 bolts holding the side covers and the output flanges are held by a snap ring just above the splines. You can easily pop out the flanges with a screw driver. These diffs nearly all use the 10mm bolts for the CV joints. These diffs are also narrower than the earlier diffs. If you want to put these into a 2002 you should use spacers to bring them to the proper width. Inner CV joints from a 1979 (and only a 1979) e21 are thicker and can be used instead of spacers if you happen to come across a set. Otherwise, billet aluminum spacers and the proper bolts are advised.
Making the Swap:
It's common to put e21 diffs into 2002's either to get the 3.90 ratio or the easily available 320is limited slip. The e21 diffs will bolt right onto the subframe and driveshaft. Just remove the 320i rear cover and install your 2002 cover. Please note the possible difference in widths I mentioned above.
The final issue is 8mm bolts vs. 10mm bolts. If your new diff takes 10mm bolts for the CV joints, you have a couple of choices. Put 320i CV joints on the inboard end of the 2002 axle (if you can find 1979 ones, so much the better), or option two is to drill and tap six new 8mm bolt holes in the e21 output flanges. Clearly these need to be located with some precision and thus we offer the service at Ireland Engineering.
How To Build the Strongest Small-case Limited-slip Differential For Your 2002.
AKA: Late E36 318i/ti Pumpkin into an E21 Open 6-bolt Differential Case.
By Andrew Adams of Ireland Engineering
Email: a n d r e w a d a m s @ I e m o t o r s p o r t . c o m
Website: bmw2002.com or iemotorsport.com
Disclaimer: This is not a beginner project, nor is it meant to be a complete diff-bible, and as such, I (or IE) are not responsible for ANY of your screw-ups. There are some topics which are abbreviated both intentionally and unintentionally.
This would not have been possible without the constant teaching from Jeff and Jeremy of IE and knowledge of Rob at Precision Gearing. If this project looks a bit much for you, or you have the money but not the time, call Rob (http://precisiongearing.com/ ).
TABLE OF CONTENTS V6.5----------------------------------
– Basics: How a differential works
– Basics: BMW Differentials
Why the E36 318ti LSD?
Different Differential Cases
Rear Cover Options
– Basics: Performance Options
– Differential Rebuild
- Tech photo w/ part names
- Test break-away
- Removing diff from car
- Side-cover and LSD removal
- Pinion Shaft removal
- Pumpkin Break-down and reassembly
- Pinion Shaft installation
- Pumpkin insertion
- Side-covers and Setting Lash
Basics: How a differential works.----------------------------
For those who'd like to understand the principles of the differential itself, I've attached some videos that explain it far better than I could (thanks Chevy and Toyota!)
Here's a link describing the basics of a clutch-type limited-slip differential (which is what we'll be messing with).
Basics: BMW Differentials.------------------------------------
BMW uses three different sized differentials. SMALL (168mm ring gear), MEDIUM (188mm ring gear), and LARGE (210mm ring gear).
For the sake of this write-up we'll be focusing on the small case limited slip differentials. These are found (as standard or option) on 4-cylinder BMW's in the 2002, E21, E30, and E36 cars.
Why the E36 318ti LSD?-----------------------------------
The E36 318ti lsd's were the last of the line for smalll case lsd's, and as such were the most capable in handling higher amounts of torque than earlier diffs. This is evidenced in a couple of ways.
– The lsd-cap is .5” thick, the thickest out of any bmw/ZF 168mm diff.
– The output shaft collars are noticeably thicker than the E21 & E30 pieces. It is my understanding that they are the same size as the pieces on the the medium-case differentials, but I have not verified that.
– The ring-gear bolts have a serrated underside for better grip.
These address the most common failure-points on the small case diffs, the cap cracking (or the bearing nose shearing), the splines on the spider-gear shafts shearing, and arguably the most common failure point; the ring gears bolts backing themselves out. With this center section I will have the best chance of having a small-case diff that won't blow up when powered by the 6-cylinder M20. This means there is no need to resort to modifying the rear subframe (or paying someone else an obscene amount of money) to fit a medium case differential.
-2002 came with 4-bolt side-covers and bolt-in output flanges (with the ’76s having slightly longer output flanges).
-E21 EARLY (08/79 and earlier) – 4-bolt side-covers and circlip-held output flanges
-E21 LATE (09/79 and later) – 6-bolt side-covers and snap ring held output flanges
I’ll be using the late E21 open case for this with the beefier 6-bolt side-covers and 10mm cv-joint bolts (in the vain hope it will matter). You “should” be able to use a 4-bolt open 2002/e21 case if the center section coming out of the case is the same width as the Limited-slip pumpkin.
Rear Cover options on the 2002 with small case differential.----------------------------------
-2002 EARLY (mid '74 and earlier)
-2002 LATE ‘finned’
-Alpina Chinese knock-off*
-Big Billet Cover*
-Tap in AN fittings and run a pump/cooler.*
This differential will be using the 02 late finned cover, all the fancy options would be overkill on a street car. I want this differential to be well-engineered, not over-engineered.
The shorter the gear ratio (higher number [4.10,4.40, etc.]) the quicker the acceleration, but with a reduction of your top speed. There are many differential calculators online that should help you to find your ideal setup.
Basics: Performance Options-----------------------------------------
Rather than go into this in detail, I'll be straight forward. Unless this is a dedicated track/rally/autox car then the extra money spent is largely wasted. The only thing I'll be doing is increasing the brake-away torque, which I will not elaborate on but you should definitely consider.
If you do have a race-car and could benefit from different ramp-angles, added clutches, polished gears, safety-wire, etc. Then please call a 2002 performance parts supplier you trust (do NOT just use anybody and skimp) if you'd like some recommendations then feel free to email me.
1 Exploded view of the differential case and its components.-------------------------------
2 Remove the differential from car (see Marshall's excellent writeup ----------------------here ….
3 Break away:-------------------------------------------------------
Prior to disassembly, measure the break-away torque of your differential (how much force it takes before the LSD clutches slip). A good percentage of the used E21 LSD's out there today are worn to the point where they hold barely more than a sticky open diff (sorry for those of you who paid $400 for an non-inspected E21 LSD). If you can twist the differential out-put flanges in the opposite direction with your hands then you've got a very worn differential. A healthy stock e21/e30/e36 small case LSD should take between 25-30 ft/lbs before slipping. By measuring now, it will give you a good baseline indication on the condition of the individual components inside the center pumpkin.
I measured my E36 pumpkin to have a break-away of 26ft/lbs, so technically I would be ok if I was to just stick it into the case and call it a day. However, I'm shooting for a bit more than 40ft/lbs. which is about the most I'd want out of a street-oriented LSD. (by comparison SPECE30 medium case LSD’s are limited to 65ft/lbs., and most other 2002 race setups are around 80ft/lbs.)
4 Side-cover and LSD pumpkin removal:-------------------------------
First take off the rear cover and drain the oil, this stuff has a rather distinctive smell. Be sure to do this in a well vented area, or near your mother-in-law.
Pop out the diff flanges, this might take something rather long (big flat-head screw-driver. crow bar, or bar stock.). To prevent damage to the side-cover, use the bolt heads for leverage. With the flanges out, use a standard seal puller to remove the flange seal out of the cover.
Remove the side cover bolts (loosen progressively in a cross-hatch pattern so as not to warp the cover [think head bolts/studs]). With the bolts out gently tap around the side-cover with a thin flat-head screw-driver acting as a wedge; if care is taken you will not leave any lasting marks. As the cover starts to come off you'll see some thin shims, these are important to keep with their respective side as they allow for correct laterally spacing of the center-section inside the case. When you pull the side covers off the center-section will come loose and drop in the case, put some rags underneath to pad the fall or risk chipping the ring gear.
String the side cover and shims together while marking the which side of the case they came off of (I simply used a die punch marking one or two corresponding dots) so when it goes back together everything will be in order.
If you are swapping in a LSD pumpkin, remove it front it's case in the same manner.
4) Pinion Shaft removal:---------------------------------------
Now that the pumpkin and side-covers are removed you're left with the Pinion Shaft and the case itself. To start you'll need to pop-out the old lock ring on the backside of the input flange, you can remove it with a flathead screwdriver or drill.
Now you'll need to loosen the nut and to do that you will need to keep the shaft from spinning. BMW made a special tool that looks like this (if you have one, chances are you don't really need any instruction on how to do any of this). You can easily replicate something to imitate this tool's function. Loosen/remove the nut and tap “in” the shaft from the backside to remove the pinion-shaft. DO NOT take a typical hammer to the backside of the shaft; you will damage it in some way. Use a rubber hammer, and if that proves difficult, use a press.
With the shaft removed you will see that the pinion-shaft has the inner roller bearing pressed on it Toss the crush sleeve on the pinion-shaft shaft into the trash. To remove the inner roller bearing from the shaft you'll need to use the press and a strong bearing splitter. If it proves difficult (like it did for me) then use a torch to heat up the bearing a bit.
Now let’s go back to the case and focus on the inner and outer bearing races left in the case from the shaft we just removed. Start by removing the circular baffle with a couple good punches with the flathead.
Punch out the inner race (preferably with a softer metal rod like aluminum to keep your punch from gouging the insides of the case.)
As the race comes out you will see a spacer behind it DO NOT LOSE THIS, it spaces the pinion gear's depth in relation to the ring gear. Each spacer is specific to that particular ring and pinion.
Punch out the outer bearing race from the inside
You can also use a clawed bearing puller but I found the punch method much easier.
With this done the case is now fully disassembled into its sub-components.
With the case completely apart clean, everything thoroughly. Thoroughly degrease the inside of the case; you don't want any muck causing havoc on your new bearings. Ideally, bath the case in a solution of degreaser. You can take some emery cloth (or scotch-brite pad) and rub down the pinion shaft along with getting all the teeth of the ring gear and pinion head. Then use brake cleaner (or otherwise) to make sure no residue is left behind. Any spec of grit can damage of your refurbished diff.
Prior to paint I broke out some die punches and stamped the case IE/AA/001 3.64.
Pic shows post paint.
Prep and paint the case in the paint of your choosing (engine enamel, por-15, etc.) likewise, with the rear diff-flange and side-covers.
For giggles I polished the rear cover, do whatever makes you happy (understand that the rear cover helps to dissipate heat so powder coating, painting, or otherwise adding insulation is probably not the best idea in the world.)
LSD PUMPKIN BREAK-DOWN & REBUILD-----------------------------------
First I compare both the E21 open center-section and the E36 pumpkin, they are the same height, and therefor compatible with one-another.
You'll notice the E36ti pumpkin has a pinwheel on one end, this pinwheel is read by a sensor on the diff case to calculate speed. Since we're not running this sensor, you can remove the speedo wheel (I used a die grinder).
If you are changing the gear ratio, remove the ring-gear from the pumpkin. To do this loosen the bolts a little at a time in a crisscross manner (just like head studs/bolts). With all of the bolts removed you will need to use a punch through the empty bolts holes and tap the ring-gear off (be sure to have some rags down for the ring to land on.
Now we tackle the case, using a bearing puller (or in my case, a bearing splitter and bench press) pull off the roller bearings from either end.
Using the same crisscross method as the ring gear bolts loosen and remove the allan-head bolts attaching top cover to the case. As you loosen the bolts you'll notice the top cover is held in tension by the as the internals expand. (Note the picture still shows the bearing in place, just ignore it)A weak limited-slip (with a low break-away) is simply due to a lack of tension on the internal components i.e. low break-away = worn internal parts = lost material = more space in case = less tension under load.
BLOW UP DIAGRAM OF INTERNAL COMPONENTS--------------------------------------
As you remove the cap you will notice 2 smaller washers (#2 & #3), the top piece is the smaller Diaphragm Spring aka Bellview Washer (make sure you see what way direction this goes in, you don’t want to put it back in upside down). The second washer is the Stop Disk, it has two indents on the bottom that allow for oil flow.
The first large disk is the larger Bellview Washer (#4), this provides much of the tension in the differential. Inspect the top surface near the inside diameter, a worn Washer will have a wear ridge there, if you can feel a change in height with your finger nail, replace the washer.
Next you have the dog-eared Outer Disk (#5), this disk acts like the surface face of a flywheel. It's available in a few thicknesses, with the thickest giving a higher break-away (but go too thick and you'll essentially have a locked diff. with no break-away). Check for any grooves or ridges and replace as necessary. Just as a flywheel you can also “resurface” the side of the plate which faces the clutch plate, just take fine-grade sandpaper (500grit) and scuff the surface evenly. The Outer-Disks on this diff looked and felt smooth, so all I needed to do was a little scuffing.
After the dog-eared Outer Disk you have the Inner Disk aka clutch-disk (#6)(and as you'd imagine it functions in the same way the clutch-disk works on the transmission). These clutch-disks have a grippy high-friction material coating on them. If your disk has no or little material left (will feel smooth to the touch) then it will slip much easier, replace it. These can develop a smooth inner ring and still have a grippy outer-ring, if this is the case, still replace it.
At the center of the LSD we have the Spider Gears and their housings.(#7, #8, & #9) The spider-gear housing pieces are officially known as the Compression Rings (#7) since they are the pieces that moves out-ward under load which put pressure on the clutch disk “locking” the differential. Like the dog-eared outer disk you can lightly scuff the surface.
Just inside and poking out of the Compression Ring, we have the Side Gear (#8) portion of the Spider Gears. The output flanges stick into these so the beefier the better. Check for cracks, if you can, get them magnafluxed along with the other parts of the spider gear and Ring & Pinion.
Finally in the middle are the 4 Bevel Gears and shafts (#9). Check them over carefully for hairline cracks.
From that point on everything on the bottom half of the case is a mirror image of the top, inspect everything and replace as necessary. Once the guts are all ready, use 2 new allan-head bolts [8x16 10.9] to secure the diff-top. Using a press, install the new roller bearings on either end. Make sure to only put pressure on the bearing's inner sleeve. Ignore that all the allan heads are in place in the pictures.
The last thing to do on the pumpkin is to bolt the ring-gear in place. First make absolutely sure there is nothing on either the ring-gear or differential mating surfaces (clean with denatured alcohol or similar). The ring is a snug fit so you might need to warm it up a bit with a torch before sliding into place (Hopefully you’ve waited for the denatured alcohol to evaporate, otherwise torch and alcohol equal BOOM!). Use some longer 12x1.5mn bolts to thread into the ring gear to act as guides as it’s lowered in place. With new bolts (or safety wired original bolts) torque the ring gear bolts down to 110ft/lbs using the slow crisscross method. When they are torqued down apply a dab of paint-pen paint to each bolt edge, this will allow for visual check when you want to make sure the bolts aren’t backing out. (If the paint seal is broke then the bolt has moved.)
Bolt down the top-cap with a couple of the allan-heads (not final torque), and check the break-away torque now while it is easy. I measured mine out to 43ft/lbs before slipping; perfect!
PINION GEAR INSTALL---------------------------------------------
Tackle the pinion gear assembly first, and the first step is to install the new inner and outer races. These can be rather tight, so it's a good idea to have them sit in the freezer a few hours prior to inserting them. On the inner race be sure to set the spacer/washer from the disassembly in prior to tapping the race in. Start tapping the race in until it sits snug on the washer/spacer (tap in the crisscross method to keep the race from going in sideways).
Use a press to install the new inner roller bearing onto the pinion shaft.
Slide the pinion from the inside, then from the other side slide on a new crush sleeve, roller bearing, seal, rear output flange, and the nut. Get the nut started by hand, but then attach the BMW special tool (or your equivalent) to the rear output flange and keep tightening with the socket (though you might need to switch to a breaker bar). At this point it's extremely helpful to have a couple of friends on deck to keep the diff from flopping over.
The factory manual calls for 25 In/lbs. of rotating torque. As you tighten the Pinion Shaft Nut down you'll feel quite a bit of resistance from the crush sleeve, be sure to stop frequently to rotate the flange/pinion by hand. With every twist the pinion should spin easier. Repeat this process until you hit the 25 In/lb mark continually (that is to say the pinion doesn't loosen as you twist) the final feel is similar to putting a finger in molasses (thick but smooth).
mlytle edit - the below is TobyB's method. I have used this way to do the pinion TQ also. several manuals have 25 Ft/Lbs as the tq, but it should be In/Lbs. even that is too high. should do it by feel like a wheel bearing as TobyB describes below.
Yeah, it's inch- lbs, that's a misprint.
When you're doing it, you'll find that once you start crushing the sleeve, you're pulling REALLY hard to crush it.
So you'll cautiously creep up on it- tighten a little, wiggle, tighten, wiggle--- when it gets to the point where you
can ALMOST not wiggle, you have to go very slowly- one tighten, all of a sudden, the wiggle's gone, but it turns very easily.
THEN you need to turn about 3 degrees more to where you can feel 'some' bearing drag. If you've set up tapered roller
bearings before, you'll know it- it'll go 'firm' but not crunchy. If you go 2 degrees too far, it'll get 'stiff' or 'crunchy'- and you
get to do it again, with a new sleeve and probably seal.
The first sleeve is easy to kill- after that, it's pretty easy to do a good job.
With the nut torqued down, tap in your new lock-plate (it will be tight but it will expand into the relief slot on the flange [if you don't bend it in half]).
SIDE-COVERS & PUMPKIN INSERTION --------------------------------------
One of the more difficult parts of the entire process is the removal of the side-cover bearing races. Typically a race will have some sort of lip to use a punch or puller on allowing for removal, for some reason the E21 open diff's side-covers have no such access.
The only way to remove them is to carefully drill a few small angled access holes on the face of the cover so you can run a punch through and pop out the race.
(Then mask one end of the hole and fill with some JB weld). If there's a better way, I'm all ears. EDIT: In the comments section below there are a few other options that have been added!
Put the side-cover bearing races in the freezer for a half an hour (shrinking them slightly just as before with the pinion).
With the JB weld now dry and the cover's race surfaces clean and flush, take the new race out of the freezer and using your old race as a driver, tap the race in as before. With both side-covers races in, remove each of the old rubber O-rings and replace them with new ones. Press in new out-put flange seals.
With your side covers (and their respective shims) ready it's now time to address how to get the E36ti assembly in without any modification. If you try and take the pumpkin and simply slide it in you'll see that it will not fit.
Remove the 2 Allan-head bolts you bolted in to secure the pumpkin top-cover. Take care that you don't spill the pumpkin guts. With the top-cover removed there is now plenty of room the slide the pumpkin inside the case.
With the case inside, slide in the top cover and simply start bolting the Pumpkin top-cover through the side-cover opening. Use new bolts and some loc-tite,
torque the bolts to 25 ft/lbs using the crisscross method, to get that much torque you'll need to slide the passenger side flange in through the case and mount it stationary (bench vice), this will allow you to tighten a bold, back back the flange off, and rotate the pumpkin until all of the bolts are torqued.
Now with the pumpkin ready, slide one side-cover/shim on (use a little assembly grease) and tighten down 2 bolts. Tip the diff onto the side that has the side-cover bolted in and maneuver the pumpkin so one end sits in the side-cover race, this will line up the diff so you can slide in the other side-cover from the top, tighten 2 bolts as before. Now you're ready to start testing for lash.
Please note that the factory manual goes over setting correct lash in-depth using precise measurements. After doing a few diffs via the manual you can start to feel the differences by hand. This is very hard to convey via a technical write-up, but I’ve given it a shot.
Lash defines the contact between the ring-gear and the pinion-gear. Too tight and the friction between the gears will be enough that the teeth will start flaking metal and stress-cracks will soon brake the gears. A too-tight lash can be identified by having a contact patch on the inner portion of the ring gear teeth (and there will be NO give when turning the input flange). Too loose and the gear teeth will shear clean off during high bursts of torque (i.e. drag racing, donuts, burnouts, etc.). This is identified by having a contact patch on the outside edge of the ring gear teeth. A proper lash will have a full contact patch on the middle of the teeth and a little give before the gears mesh which can be felt when turning the input flange.
To adjust the distance between the gears (thus adjusting lash) we use the side-cover shims
These shims compensate for manufacturing tolerances of the pumpkin and can move it laterally within the case. A thicker shim on the driver's side will set the ring-gear further away from the pinion-gear. A thicker shim on the passenger side pushes the ring-gear closer to the pinion-gear. I used Dicum layout fluid to observe the gear contact patch, you can use any sort of marking paste or fluid.
Many shim sizes are still available through BMW and typically cost $8-$10 each.
With all that in mind, the total shim thickness for the two shims on my e21 case was.120”
I test fit the new pumpkin with each shim as it was originally. This proved to be way too tight, there was no give in the input flange when turning and a test swatch of Dicum on the teeth showed a contact pattern on the far inside. For the next test I stacked both shims on the driver's side (pulling the ring-gear away from the pinion gear), that resulted in too much lash (about 1/4” inch before the gears contacted one-another) but it was better than the first test. My ideal lash was somewhere in between, going onto ETK and converting metric to standard I ordered the thickest shim available (.08”) and a corresponding (.04”) shim to maintain the .120” total thickness.
Once the new shims came in I put them on and broke out the Dicum once again.
In the initial twisting of the input flange, I found there was about 1/16th” lash before gear contact (near perfect!). Rotating the ring-gear through yielded a nice fat contact patch in the middle of the ring-gear.
Lash set, and ready to go.
For the final assembly of the side-covers I pulled the 2 test bolts from either cover, added a touch of loc-tite, and torqued down all 6 bolts for each side. As the Piece de resistance I popped on the polished re-drilled output flanges. I'll be using the stock '02 8mm cv-joints for now but will switch the the 10mm E21 cv-joints once the M20 goes in.
Lastly was the rear cover gasket and rear cover. With those in place I had a complete, tough, limited-slip diff ready to be installed and broken in.
Had not seen full “how to” for diff swap, so here you go.
Simple job. Well, simple on a car that gets its diff swapped a lot. No promises on a diff that has been in a car for 40 years…
What you need:
Two 17mm combo wrenches
19mm combo wrench
6mm hex socket (for 2002 cv joints)
8mm hex socket (if you have 320 cv joints)
First, get the car up on jackstands.
Lock e-brake and put car in gear
Loosen the 4 nuts connecting drive shaft to diff (17mm). you will need to get at them from the right side of diff and will have to rotate the drive shaft and re-lock once.
Loosen the 24 hex bolts that hold the half shafts to the diff and wheel hubs. 6mm hex for 2002 cv joints, 8mm hex for 320i joints. You could get away with just doing the inner ones, but gives you more room to work if you just take them all the way out.
Creative use of a combo wrench applied to a short hex key if that is all you have.
Or use hex socket and drive
Halfshafts out. Good time to check cv joints and refresh if needed.
Now get a jack under the subframe to support it. without the diff attached, the subframe will sag. Not good for the mounts.
Remove hanger bracket. I just a tranny jack under the diff to support. Or you can have a buddy bench press it.
17mm combo wrenches on the bracket, 17mm socket on the nuts holding diff to bracket.
remove the 4 bolts holding the diff to the subframe. 19mm combo wrench on top, 19mm socket on bottom. impact gun is wonderfull thing here.
Lower back of diff first. When clear of tank, slide out.
If swapping in a 320 3.91, you will need to use the 2002 diff cover. Good time to check the innards and reseal the cover anyway.
Remove cover with 17mm socket and scrape the old gasket/goo off cover and diff.
Bmw specifies a paper gasket on the cover, but I have never used one in the dozens of diff covers I have sealed. Never had a leak. I use permatex ultra black sealant instead.
Apply a bead around diff
Replace cover gently. Screw bolts in until just snug by hand. You should see just a little ultra black ooze out. Let diff sit for an hour or two for goo to set.
Once goo has set, torque cover to 32ftlbs.
Get new crush seals for fill and drain plugs….these are on use items. Do not reuse.
Now is the good time to refill the diff with oil. I use redline 75w90 for street diffs. Redlilne lightweight shockproof oil for track diffs.
Assembly is reverse of the above process…..but when I put mine back together I will add detail to this.
much easier to fill diff like this than when it is in the car. takes 1.1 qts
Take lots of pictures, everything, every angle.
Disconnect everything, wires, hoses, distributor cap. You may want to remove the front grills and radiator but I don't think that was necessary.
Leave the rear wheels on the ground, well chocked.
Raise the front end.
Front of car secure on jack stands.
Remove the front wheels.
Remove the bracket on the strut that holds the rigid brake lines
Remove the brake calipers and hang them from something with something.
Get the car high enough to get a jack under the front sub-frame. A floor jack or an ATV/motorcycle jack works fine.
Disconnect all the shifter/transmission stuff from the tunnel.
Disconnect any and all electrical wiring
Disconnect fuel lines
Disconnect the exhaust at the downpipe
Disconnect the strut bearings from the inner fenders.
Now, with the sub-frame/motor balanced,lower it slowly, with a friend watching, to make sure everything clears the frame rails. The shifter and distributor tend to be the things that hang up.
Once the motor clears and is as low as the jack allows, move your attention to the floor jacks and start moving the car up until the motor will clear the front valance. Pull it out, lower the car.
Watch the struts, they want to fall outward. Have someone hold them or bungee cord them to something.
Mating transmission to motor the hillbilly way
Motor plant today with the help of Bill Riblett
Pick a nice cold day to work.....
Mate the transmission and motor
Mate this to the sub-frame
Balance sub-frame on jack
Lift car far enough for the motor, sub-frame, jack to go under the front valance.
Move the motor under the car. Eyeballing it to make sure everything will clear, line up the shifter.
Lower the car as far as you can, slowly, having a friend to watch everything. The car should either go so slowly you can hardly see it move or lower it a bit at a time, stop, look, lower, stop, look, you get the picture.
Now raise the motor until the sub-frame mates to the frame rails. (Watch and align the strut studs through the inner fenders.) Use a big punch, 10incher, to line up the holes in the frame rail and the sub-frame. Secure the bolts and tighten. If these fight you, get longer bolts of the same diameter and pitch. Use the longer bolts to get things lined up. Pull the sub-frame and the frame rail together, when you can, change out the longer bolts for the properly sized bolts
Secure the shifter stuff and transmission mount and support. Make sure the speedo cable is inserted before the transmission support it put into place.
Reattach the brake line brackets and the calipers.
Original Author: Paul Wegweiser
Removing and Rebuilding the Pedal Box
Pedal Box rebuild kit, consisting of: 35 41 1 108 237 Throttle bushings x 2 35 41 1 108 676 Throttle lever spring 35 41 4 440 122 Gas pedal roller 35 31 4 640 116 Clutch pedal to Master bushings x 2 35 21 4 640 103 Clutch pedal pivot bushing 35 21 4 045 707 Brake pedal pivot bushing 35 31 1 104 531 Clutch pedal return spring 35 21 1 102 383 Brake pedal Return spring 35 21 4 440 113 Clutch and brake pedal pads x 2 (THANKS: scottt in New Hampshire for the list of part numbers!)
Removing the Pedal Box has to be one of the most-dreaded operations one can perform on an '02. Probably only pulling the heater box is worse in terms of popular perception of the difficulties involved! Total time required for the operation is maybe 3 hours at most - and a truly rewarding repair!
The first time I performed this job with the motor OUT of the car - and it was still a royal PITA! I must've wasted 3 hours on a couple circlips and link rod pins. So here's the secret magical method for painlessly removing and installing the dreaded pedal box from an '02.
First off: You don't need to disassemble as much as you think you do! THAT is the secret. Pedal box bushing and spring kits are about $35.00 or so - and worth every penny!
1. Peel back carpet so you can easily reach all the pedals, and pedal box bolts. If you've got a later 74-76 car with the one piece carpet - be prepared to go submarining under that nasty funk filled carpet with a flashlight in each nostril and wrenches in your mouth.
2. Loosen the large bolt that holds the clutch pedal to the master cylinder push rod. It's a 19mm / 13mm thing - and it SHOULD have bushings on either side. These generally crack or decompose during removal - glad you got new ones, right? Now pop off the tension spring and circlip holding the vertical rod and clevis block to the brake pedal - push pin off and through toward the left to disconnect. Sometimes this pin is stubborn - but trust me - it'll move out of the way. Don't mess with threading the clevis block unless you want to adjust pedal height. We'll save that for another sermon. Remove all the 13mm bolts that hold the bucket/box to the floor.
3. From the engine bay side - follow the pedal box as it travels up the firewall - and you'll notice it's held in place by a long fat bolt that passes through the brake booster bracket. You do NOT have to remove the bolt to remove the box (that's secret number TWO!) Since the box is slotted at the top - just loosening this nut/bolt to it's last possible thread will allow the pedal box to drop down for removal. If you remove that bolt - you'll spend the rest of the day wrestling with pivot arms, mean, miserable hidden circlips, and scaring away small garden animals as you hurl tools about the yard in primal rage.
4. Now to get under the car. If you were paying attention to the gas pedal as you popped it off, you'll see that there's a steel rod that passes through the side of the pedal box. Now's your chance to start its removal process. Loosen the 10mm nut/bolt that holds the arm lever to the pedal rod, make a note of the spring orientation, and hop back inside the car. Now - from inside the car - wiggle the pedal rod while you pull it to the LEFT - and out of the pedal box. It is NOT possible to remove the box without removing this gas pedal rod first. Trust me - I've tried it six ways from Sunday.
5. Now the hardest part of the job - removing the clutch master cylinder from the box. Reaching those 13mm nuts is a true pain. But if you remove it from the box carefully - you won't have to unhook the supply line or bleed the clutch at the end of this little trip. Plus - a loose supply hose dangling above your head is gauranteed to dribble brake fluid on your noggin! I don't know about you, but I hate brake fluid more than any thing in this world. As if it wasn't bad enough that it stings in cuts, blinds you, makes everything slippery, dissolves latex gloves and stinks; but it'll take the PAINT right off your car too- YAEECHHK!!! Horrible stuff! Treat with care!
Now there's only one obstacle holding the pedal box in: the metal brake lines where they are clamped to its side. CAREFULLY pry the tabs back, pull the lines away from the box bracket, and lower the box down. I work from under the car; guiding the clutch and brake pedals down through the hole in the firewall.
6. Once it's out - take a GOOD look at the arrangement of the clutch spring, washers, and bolt holding the pedals on. remove bolt, replace all wearable parts (springs, bushings, circlips) making sure to lubricate everything with a good (I'm a 'Moly guy myself) grease.
7. Reinstall box, clutch master (you'll cuss at the master cyl. It's OK - we all do it!) and finally the gas pedal rod. Having a buddy (thanks Tim!) push the pedal to the floor while you slide on the splined throttle shaft arm at full open throttle (at the carb) will greatly speed up the operation.
Best O' luck to ya'! The new feel of the pedals will amaze you! Truly worth all the effort - and only required once every few decades!
If you have any questions please post them on the Message Board!
Original Author: Sam Schultz
Valve Adjustment for BMW 1602-2002 M10 Motor
The process of adjusting the valves needs to be done at every major service. This service happens every 15,000 miles or so. Failure to properly adjust the valves will cause increased wear and rob your engine of power.
10-mm socket (optional)
0.008" feeler gauge
Small allen wrench or piece of bent coathanger
Valve cover gasket (11-12-1-734-276)
Project: With the engine cool (let it sit overnight), remove the valve cover, 6 castle nuts and 1 bolt. Remove valve cover and, depending on its condition, you should replace the old valve cover gasket as it is NOT reusable.
You can start at any cylinder you like, but just keep track of where you start! What you want to do is relieve all tension on the valve you are adjusting by rotating the motor until the valve is fully seated. If you are using the starter to rotate the engine make sure the car is out of gear, e-brake engaged and the positive coil wire is disconnected. If you are rotating the motor by the crank nut remove the spark plugs, and follow the same set up procedures as with a starter switch. My favorite way is to simply roll the car while its in gear.
To do this, use a flat spot of hard ground. Disconnect the + coil wire and put the car in 4th gear. (Optionally, you can remove all the spark plugs to make this slightly easier-- you should inspect them now anyway.) Roll the car slowly until the first rocker arm you are doing is on the backside of the cam lobe. Next, tilt the rocker pad forward until it is resting on the camshaft. Back off the lock nut slightly and slide the feeler gauge under the eccentric.
Use the small allen key or bent piece of coat hanger to rotate the eccentric until its snug on the .008" feeler gauge. Then tighten the nut. The nut should be tight but don't break or strip it! The gauge should just slide between the eccentric and the top of the valve and should grab slightly as you remove it. Make sure you hold the feeler gauge flat. Be aware that the valve spring itself may compress slightly if you turn the eccentric too much, causing the valves to be too tight when you tighten the eccentric. It is better to have the valve a little loose than tight, but if it is too loose it will cause lower performance, valve clatter and wear. Too tight will also rob power by preventing the valves from closing all the way, and also wear out the cam/rockers/etc. much more quickly, and THEN cause valve clatter...
The entire operation can be a little tricky but a novice mechanic can do the job! If you are good you can do an intake and an exhaust valve at the same time (not the same cylinder). Make sure you adjust them all, then replace the valve cover gasket and reinstall the cover. The acorn nuts should only be torqued to 8lb.ft., so don't overtighten them! This will warp the cover and cause leaks, assuming you get away with not breaking the studs themselves. Once you are finished, start the car and listen for any strange noises. Remember vintage BMWs should have some valve noise, not having enough play will cause wear to the cam shaft. The old rule is its better a hair loose than a hair tight. If all is good, then close the hood and take it for a nice run!
Original Author: pklym
Well, I am trying to slowly prepare my '76 02 for a late summer cross country trip. Three years ago I brought the car from Portland Oregon out to Washington DC to attend law school. With six weeks before finals I am finally getting serious about preparing the car the trip back west (will probably leave the car stored somewhere in DC area for the summer before driving it out in August if anyone has suggestions on a safe place to store the car out here). I struggled through the winter months rebuilding my heater box...a few times until it worked right but now that we just had a good rain to wipe the streets clear of snow, I am getting to some more preventative maintenance.
After reading everything I could find on adjusting the valves, I decided I would tackle this job. The valves didn't sound particularly off, but my engine does feel a little tired, so I thought I would give it a try. I'm not sure if I have ever had the valve cover off of Bridgette and have no idea when they were last adjusted (I've owned the car for four years, but other than the cross-country trip, and one from DC to Birmingham, AL, I haven't put too many miles on it). I read as much as I could but still did not completely understand the process, mostly I thought "What the hell is an eccentric?" I could not find any pictures of that. Because of my slight confusion, I thought I would try to document my process to make it easier on other first-timers.
First things first. Remove the spark plugs, and wires (and I removed the coil wire, I think this is redundant).
Then I removed the valve cover breather hose (which is vented straight out the bottom, at some point I will get a catch can and maybe reroute it back into the carb). I undid the VC bolts a little at a time working from inside out, I am not sure if this is necessary but I didn't want to accidentally warp anything.
Now I tried to get a lay of the land, I understand the cam no problem, so I looked for the lobes that were pointing straight down (as oriented to the engine, not the ground) to see which valves to mess with first. It is kind of hard to tell when the valve is completely straight down because you can't see down there, but you can tell when the lobe is not touching the rocker anymore.
After locating a valve to work on, I slid my .06/.08 go-no-go feeler gauge into the spot just above the valve. I must admit, this took me a bit of time to figure out exactly where it went, but after some trial and error it made sense. I had to bend the gauge pretty severely to work in the area.
From there I had to figure out what this eccentric thing is. In the picture below I labeled the nut you have to loosen (it's pretty tight and will move the rocker at first) with a "1" and the eccentric with a "2"
For the most part the valves seemed in good adjustment, none of them fit any feeler in higher than .08 so they did not need much adjusting. I felt like it was tough to get the resistance on the feeler exactly where I wanted, but ultimately I just relied on the go-no-go aspect of the feeler.
I was confused at how to adjust the eccentric. I knew you could use either a small allen wrench or a coat hanger, but I also read that you turn in the hole of the eccentric. I thought this meant you actually rotated some sort of nut. That's not so, you turn the eccentric along its pivot point. I marked the photo with the way the eccentric moves.
I also quickly learned that you have to keep your eccentric tool in the eccentric hole while tightening up the nut or the eccentric would close the valve gap too much. I had to switch to a piece of coat hanger from my original allen wrench tool to get a few of the eccentrics to turn properly. I snugged up the nut, but didn't tighten it too much. I hope that is right (anyone?).
After doing the two valves whose cam lobes were pointing down I got in the driver's seat and bumped the starter just a bit and would check what valves were available next, without a helper I had to head in and out of the car a few times to get the cams just right. I did try turning the motor with the fan and the alternator belt, but I could only get it to turn a bit this way, I think I may need to tighten my belt.
After checking and adjusting all the valves (a few of them didn't need any adjusting). I put a new valve cover gasket on with a couple dabs of Indian Head Shellac in the corners. I torqued the VC bolts to 96 inch lbs of torque (8 ft lbs), working inside out. This really isn't much. Gapped and inserted new spark plugs, and hooked everything back up.
Thankfully the car started back up and seems to run well. Not sure if there was any improvement, but then again they didn't seem much out of whack. I haven't taken it for a spin yet.
Next, I will be replacing all my fuel line (except the in-car hose) and sender gasket. I think my carb leaks a bit of fuel onto the intake manifold which ends up making my car hard to start without starter fluid if it has been sitting for more than a few days, unfortunately I don't think I am going to have the time to try and learn how to rebuild a carb before my trip. I would have done the fuel line today but the lady at Autozone (God I hate Autozone but its my only option nearby) threw some 5/16" VAC line in my bag instead of fuel line and I didn't notice until later.
I hope these photos help someone, and if anyone who actually knows how to adjust valves and sees anything I did wrong, please point it out!
Original author: johnhup
I have decided to go a different route then most and install two indivual throttle bodies (ITBs) instead of a single throttle body and plenum type system. So far everything is working great based on Speed Density (Maifold pressure reading) which many thought would not be possible. I also have a more agressive cam with no overlap which probably makes the situation easier (278 duration, .288" lift). More information to come as I install EDIS spark control. Until then I wont be able to give a full review.
2 TWM 2900 DOCE 40mm's (2900 + 2910)
TWM Fuel rail
TWM pressure regulator (2800-6401)
TWM TPS switch and plug
Mangoletsi intake manifold
318 water manifold
Copper pipe bypassing manifold (to heater box)
Electronic under-car fuel pump (40+ PSI. for a Ford F150)
MegaSquirt (MS II v3. keeping distributor for now, future expansion in mind)
O2 sensor (Innovate LC-1)
12' MegaSquirt pre-labeled wiring harness
Mechanical fuel pump blockoff plate
Two Bosch temperature sensors (62110788115)
Painless wiring 7 circuit fuse box (dual relay hot switched relays)
OEM BMW intake manifold gaskets
Injectors (22# Bosch Pink Tops)
Lokar throttle cable 36"
Mangoletsi barb for brake vacuum
More band clamps (big ones for fuel pump)
Hose mounting clamps
50mm TWM air horns
EAR-100199ERL - Gauge adapter AN6 male-to-female
230206ERL x 2 - AN6 Fuel filters
AEI-12301 - AN10 10 micron fuel filter
AER-FBM2162 x 2 - AN6 to AN10 reducer male to male
EAR-809106ERL - 90 degree AN6 swivel hose fitting
EAR-310006ERL x 2 - 10ft of AN6 autoflex stainless hose
EAR-800106ERL x 5 - AN6 swivel seal hose fittings
AER-FBM1512 - AN6 to barb
EAR-360165ERL - AN6 to 5/16 hard line
RUS-650340 - Fuel pressure gauge
EAR-985006ERL - AN6 o-ring male straight cut to male AN6
The temperature sensors were ordered from my local independent BMW parts dealer. OEM part number 62110788115. They are just the simple one wire-hookup type, the housing is the ground. They were very easy to calibrate actually. A lot easier than I thought. Just get yourself a known working thermometer and freeze and boil some water measuring the Ohm's at 3 three known intervals. Here are my results:
F - Ohms
44 - 1.300
80 - 0.568
185 - 0.73
I then plugged them into my MegaSquirt and they all checked out. That simple.
Pickups (Hall, coil, VR): http://www.megamanual.com/ms2/pickups.htm
36-1 tooth, Ford EDIS ignition (future): http://www.megamanual.com/ms2/EDIS.htm
This is the story of the swap of a later transmission and front part of the propshaft into a 1967 car, to replace the broken and bent ones resulting from an exploded flexdisk/giubo.
The problems I was facing at this point included:
-shifter platform destroyed (but bent back into shape)
-front part of the propshaft bent and flange distorted
-transmission flange distorted
-mounting point for shift platform broken off the transmission
-shift rod into the transmission was also bent
I could get the original transmission to shift into four gears, but could not find reverse.
I started out by buying a second hand, later propshaft, with two universal joints and four-bolt flanges. That cost me 15 euros, and the propshaft was in fantastic condition - good central bearing, no play in the joints. I lubricated them with fresh grease, and thought I'd be good to go.
It would be rather important that the front part of the propshaft would be of the correct length, because the earlier propshaft has no universal joint but yet another giubo, the center bearing is reportedly different and no longer available (bollocks, it turns out - the dimensions are exactly the same, and the later part should just press on), and the front giubo is thinner.
These are the front parts of the shafts - later one on top, earlier one below:
As you can see, the splines are the same, and the later front part fits together perfectly with the earlier rear part.
That is crucial, because the rear part of the shaft is of a different length due to the long nose differential in the early cars, and it has a four-bolt flange as well, while the early cars have yet another giubo at the back end, with a three bolt flange.
Early rear part (note the giubo in front of the differential):
Later rear part:
Four bolt flange and UJ at the back end of the later propshaft:
Since I now had a four bolt flange at the front of the propshaft, I also needed a four bolt flange at the transmission end. Since my transmission would need work anyway after the accident, I bought a second hand one from a nice guy in Germany who had five to choose from. I picked the one which didn't leak. I could shift it manually into four gears, but couldn't find reverse ... sounds familiar ?
This is it - '72 transmission, probably a Getrag 242, came complete with clutch slave cylinder, which was a bear to remove, even after I found the second circlip.
Rear end - note the longer shift rod, this is necessary to clear the larger 8-bolt giubo.
Front end - I was worrying about the splines, but there was no indication that the clutch splines had ever changed.
Some number stamped on top of the bell housing ...
The next step was to have the friendly neighbourhood garage exchange the transmissions and clutch mechanisms.
This is what came out - nice original Getrag 232 - unfortunately damaged. Note that it still has the course splines on the output shaft. This was another reason why I could not just buy another flange - new four bolt flanges have fine splines.
Input shaft splines are identical ...
And here the number is at the back end. I couldn't find a casting date, but it should be 1967 or 1966. The car first ran on May 22nd 1967, the build date cannot be found out, as it was a CKD car assembled in Kontich. Note how the shift rod is bent and the broken off shifter platform mount.
With the new transmission, you need a new shift rod between the shifter and the shift rod at the back of the transmission. The older ones were longer, as can be seen in the pictures below. Luckily, this was included with the transmission.
Both shifter platforms, which are otherwise identical:
Newer:- yes, obviously I changed the rest of the parts over to have a shiny shifter inside the car.
Make sure to renew the shifter bushes. I've never known the car other than shifting like stirring in a very large kettle of soup, and it's been 21 years since we own it. In fact my girlfriend called me at some point to ask me where reverse was. I told her "to the left and forward" - turned out she had shifted the seat towards the front to reach the pedals, and there was so much slop in the shifter it wouldn't go into reverse with the seat that close! The two plastic bushes where the shifter connects to the shift rod disintegrate, and the rod is loose. They cost something like 2 EUR from W&N - best upgrade ever - the shifting is now tight. I tried to look for them a few years ago, but got confused by all the rings and circlips and bushes in the parts diagram, none of which seemed available. Trust me, all you need is those two tiny plastic bushes at the end of the shift rod to make your car shift like new! The location is indicated by the shiny red arrow in the picture above.
Then I bought some more new parts - apart from the bushes, I bought a new 8-hole guibo and bolts (another reason to upgrade - the old propshaft would have needed 3 6-hole giubo's at 105 EUR each - the 8 bolt giubo costs 70 EUR, and you only need one ...luckily the 6-hole one in the rear part of the old propshaft seems OK for now), a transmission support rubber, and the rubber thingies to connect the exhaust. All of those had definitively perished.
The time to reassemble came and I'm happy to report that I didn't make any mistakes that required me to undo things and start over in a different order. I started with mounting the later shifter platform onto the later transmission. This is almost impossible with the transmission in the car, while the older shifter platform can be mounted and dismounted at liberty. However, it can be done, if you rotate the shifter 180 degrees, and then you can just get it through the hole in the bottom of the car while moving the shifter platform past the (longer) shift rod from the transmission. The next step was to hook everything up and see whether the transmission shifted. That it did, like butter, also into reverse. Good!
I unbolted the transmission support beam to drop the transmission at the rear, replace the broken transmission mount and give me somewhat easier access to the flange. I started by bolting on the giubo, nuts on the transmission side.
Then I slid the front part of the propshaft into the rear part, and moved it forward until it connected with the transmission. I jacked up the rear of the car and undid the hand brake, in order to be able to rotate the propshaft and connect up the four remaining giubo bolts - again, nuts on the transmission side. I put the transmission support beam back, lowered the rear of the car, jacked up the transmission and using a bit uf muscle and a choice selection of swear words, bolted the support beam back in place. Then I bolted the propshaft center bearing back in place. Make sure it is above the support before bolting the transmission support back in place ...
This is what it looks like when done. Since the whole thing had fallen apart already when I started taking it to pieces further, I have no idea how far a correct propshaft front would protrude. It doesn't seem too dramatic like this though.
Finally, I changed the oil in the new transmission - always undo the filler plug first to avoid being stranded... What came out was gold in colour and smelled like hypoid oil - good. The magnetic plug had a substantial amount of metal deposits, but all fine and nothing gritty to the touch. Good so far.
I still have to discover a sensible way of filling these transmissions with oil - using a funnel and a long tube from the engine compartment somehow seems less than ideal, especially if the only way to know it's full is by it overflowing through the filler hole, coating the undersigned with a nice smelly layer of GL4 hypoid oil while trying to do the filler plug back up.
I had drilled two large holes in the exhaust downpipe to hammer out the dents caused by the propshaft, and bandaged those up with a tin can and exhaust bandage. Of course it still leaks - I have a new downpipe on the way now. I mounted back the exhaust manifold and downpipe (use copper grease on all the studs for future disassembly!). It turns out the exhaust studs were where all the oil over the right side of the engine was coming from. I put them back in with loctite, and hope that that will stop the leaking. Reassembling the exhaust manifold is a fiddly job, topped only by reassembling the downpipe, where the German engineers specifically designed the bolts so they can never be reached, let alone tightened, by a mere human.
Still, I managed to get it done, lowered the car off the stands, started the engine, shifted into gear, and went for a drive around the block, testing all gears - all good! Yay!
Still to do:
-remove steel band from giubo
-replace clutch spring, which went missing in action
-replace exhaust downpipe to get rid of throaty sound
-double check snugness of giubo nuts in a few 100 km
-replace and connect wire for back-up light
The result of all that can be seen in the two pictures below - new (correct, early) exhaust, which needs the exhaust support in a different position than the later system, steel band around giubo removed, back up light wires replaced and heat shrinked (in green) ...
... and finally the clutch spring:
And that's it - she's back on the road!
One issue remains, and that is the balancing of the propshaft. I've been mucking around with a hose clamp all the way up front, but that does not seem to be where the problem is located. I have the impression that the middle of the propshaft, where it is supported by the rubber-mounted central bearing, goes into a harmonic vibration at about 35 kph, in a very narrow speed range. It is probably out of balance, but the question now is whether it is the front or the rear part that is out of balance , and by how much. I'm sure a competent driveshaft shop could look at it and balance it, but that means taking the whole thing back out, which I am not looking forward to. Well, at least I know how to by now!
If you made it all the way down here, congratulations - and thanks for reading!
This is the continuation of the Megasquirt/EDIS installation article found here (which I apparently ran out of article room in):
This article will focus specifically on how to tune an EFI installation on your 2002. It will focus on Megasquirt as the example (since that is what I installed on my car and thus am most familiar with), but the general process is still applicable to other type of EFI setups.
Note: Before you ask, no, I will NOT share my actual *.msq files. That's a great way for an amateur to screw up their engine by 'just trying.' I will however share and explain screenshots of the parameters that I use in my car. It is then up to YOU to understand this and use the information to build up your own parameters. I have traditionally used Megatune but have just recently been trying out TunerStudio Lite, so you will see screenshots from both pieces of software in this write-up.
Section VII - Tuning
1.) Engine parameters
2.) Cranking and Starting
5.) Acceleration enrichment
6.) Speed-Density vs. Alpha-N and the VE table
7.) EGO and AFR table
8.) Spark advance table
9.) Datalogging and tuning
So you finally have Megasquirt (or another EFI) hardware installed on your 2002 and all wired up. But there's still more work to do before you try to crank the car up for the first time! Let's first double check that you didn't miss anything and have all of the prerequisites for loading up your first tune:
A.) All hardware is installed (manifold, throttle body, fuel rail, injectors, fuel filter, fuel pump, O2 sensor, coolant sensor, intake air temp sensor, throttle position sensor, fast idle circuit, and if applicable, EDIS hardware).
B.) All wiring is correct, hooked up, and SAFE (e.g. right size wiring, fused, etc.)
C.) Megasquirt controller boots properly, has firmware loaded, and can communicate with your tuning laptop (via Megatune, Tunerstudio, etc.)
So basically make sure everything is installed, wired properly, and you have proper communication to the EFI controller. A great first test of things in general is that when the Megasquirt unit is first powered up, it should briefly run the fuel pump for 2-3 seconds in order to prime and pressurize the fuel system. This also gives you a chance to make SURE that you don't have any fuel leaks, which obviously would be a bad thing!
I'll detail the basic setup process specific for my 2002 below, but in addition to my walk through I would strongly encourage you to also read and understand the details of the system straight from the source in the Megamanual:
There's also a great write-up on using the TunerStudio software here:
On a final note, make sure you go through and put in the base settings for each and every one of these sections BEFORE YOU ATTEMPT TO START YOU CAR FOR THE FIRST TIME!
1.) - Engine Parameters
OK, now let's start taking a closer look into the details of programming the controller, starting with entering the details of YOUR specific engine, which is what all later calculations will be based on. At its core, EFI control is actually rather simple. It just aims to inject the right amount of fuel to proportionally match the amount of oxygen entering each cylinder. But to know how much oxygen is entering each cylinder, we first need to know things like how many cylinders are on the engine, the volume of each cylinder, and how efficiently they draw in fresh air. Obviously our 2.0L engines have 4 cylinders each with a theoretical maximum volume of 0.5L of fresh air each. This is our starting point. In my Megasquirt's general settings, I have identified my ECU type as Megasquirt 2, and total engine displacement of 121 cubic inches. All other parameters I have left at their default values:
The next important part to setup is the injection control settings. This tells the controller the details of my injectors (how many, how much fuel they flow, etc.) and how I want them controlled. The first step is to click on the 'Required Fuel' box up top and enter the engine and injector details, and the program then calculates a baseline for how long the injectors need to be open to provide the right amount of fuel under ideal circumstances. The lower portion then is where you decide HOW you want this amount of fuel to be delivered. In this example for my car, I have all four injectors firing simultaneously twice per engine cycle. This means (for a 4-stroke engine) all injectors will fire once per 360 degree crank revolution, and each firing will deliver half the required amount of fuel. Other options would be to fire them just once with the full amount of fuel per 720 degrees, or every 180 degrees with 1/4 the amount of fuel. You also have the option to batch fire just one bank of injectors instead of all of them together, but I think this makes more sense V-configuration engines where the two injector banks are two separate fuel rails for each cylinder head, and personally I don't see any reason to try this on a 2002:
2.) - Cranking and Starting
In the general settings, you can choose to set starting parameters, like priming pulse, cranking pulse, after-start enrichment, etc. to use a two-point linear calculation or a table-based calculation (based on coolant temperature) that lets you do more fine-tuning of the parameters. I use the table option, but still keep things pretty linear.
The priming pulse is a bit of gas injected BEFORE you even start cranking to 'prime' the initial intake charge. I have this set from 6.0ms at beyond-cold temperatures to 2.0ms at beyond-hot temperatures:
Similarly, the cranking pulse is the injector pulse width during cranking (I think by default 'cranking' is defined as less than 300 RPM, you can adjust this but it seems fine for most cars). For this I go as high as 10ms for below-freezing conditions to 3.0ms for a fully hot-start. Obviously this is an important one for achieving easy cold starting. I admit though that I live in a rather mild climate so haven't really tested this very much on sub-freezing days, and I suspect my hot-start parameters are a bit on the rich side, but I have not fully run this one down yet. I'm basically saying, take these values with a grain of salt, and if anyone wants to propose more fully-researched values, I'd be happy to hear from you!
3.) - Warmup
But wait, there's more! After the engine catches and has started, it still wants more fuel than normal for a few seconds while everything 'gets moving,' so to speak. This is accomplished by setting an addition enrichment percentage (added to the normal VE fueling calculated value), again based on temperature. There's actually two tables involved, the first is how MUCH additional enrichment is provided (the afterstart enrichment, or ASE, percentage), and the second is for how LONG, in engine cycles, it is applied for (ASE Taper):
Ok, now we've gotten the starting and first ~30 seconds of parameters set, but what about for the rest of the time it takes for the car to get warmed up to operating temperature? Well, again there's two adjustments for this. The first is warmup enrichment. This is basically the same idea as ASE, but now base purely on coolant temperature and input as a total percentage of normal calculated fuel amount. I consider my car 'up to operating temperature' by about 170 degrees F, though you'll notice I chose phase out the warm up enrichment a little earlier at 150 degrees. Those last 20 degrees don't seem to matter much in how it runs, and for keeping the normal tuning process clean I like to not have to worry about factoring in warmup enrichment as a variable. One other variable to notice in here is the Flood Clear Threshold %. This is another starting parameter and is the throttle position at which Megasquirt assumes that you've accidentally flooded the engine and will SHUT DOWN all the injectors during cranking. This is a nice feature, but if you are used to starting your car with the throttle open, that will be a habit that you need to break! With properly set up Megasquirt tune, you should let it handle everything and always keep you foot off the pedal during starting in all conditions.
4.) - Idle
Now that we have the fuel set up for the warmup phase, obviously the next important part is the intake air. This is set in the Idle PWM duty cycle if using a PWM valve, or in Idle Steps if using an IAC stepper motor. Note that there are conditions for cranking/starting to be set here as well. For starting, the IAC can be set to a cranking position, and move to the right spot per engine temperature. This is set in the Idle Control table. The Start Value is how many steps the stepper motor retracts (opens) when first powered up. This should be enough to get it from the fully closed to fully open position. The next box, cranking position, is how far it then closes back down from the fully open position. This then is effectively how much air your letting in during cranking, or put another way, how much throttle you want during starting, but without using the gas pedal. The third important box is the crank-to-run taper time, which is the number of seconds after the engine has started that it will hold this position before closing down further to the calculated position for the current engine temperature:
Now that the starting parameters are set, next up is the warmup parameters. This is similarly adjusted as a table of IAC steps per temperature. Remember you'll want the IAC fully closed by the time the engine is warmed up, so this value in steps should match the start value from earlier, so that it always runs from fully open to fully closed:
5.) - Acceleration Enrichment
Ok, this should cover all of the basic parameters for getting the car started and idling from cold to warm, other than the main VE table. There's one last area that we want to adjust first before getting there though, and that is acceleration enrichment. Just like the accelerator pump on a carburetor, this system provides a little extra fuel during the transient periods that you are actually changing the throttle position. It can use either changes in manifold absolute pressure (MAP), or changes is throttle position (TPS) as inputs, or a combination of the two. I use a 50/50 split of both, found it's smart to raise the threshold a bit, so that normal 'noise' from either the MAP or TPS sensor doesn't trigger any unwanted acceleration enrichment to kick in. Tuning this acceleration enrichment properly will help give you that nice snappy throttle response you undoubtedly want:
You can also take this one step further if you want and try playing with the X-Tau correction tables. X-Tau corrects for an amount of fuel that gets 'stuck' on the walls of the intake port and valve (momentarily leaner), and then gets burned up *later* once it evaporates (momentarily richer), but I've never personally turned on X-Tau correction or tried to mess with it at all, so I can't offer you much advice beyond the basic operating principal here.
6.) Speed-Density vs. Alpha-N and the VE table
Finally to the main fun part, the VE table! Let's firstly understand exactly what this is. At the very beginning in the Injection Control section, we told Megasquirt how much fuel our 2.0L engine will need under IDEAL conditions with perfect cylinder filling, but what about real-world conditions? This is where the VE table comes into play. At idle with a closed throttle plate, each cylinder isn't pulling in a full 0.5L worth of air, so clearly less fuel is necessary. Cylinder filling is also different at different RPMs; even at full throttle there's a difference in how much air the engine can pull in at 2000 vs. 5000 RPM. These two variables, engine load and engine speed, are the axes of the VE table, and the values then are nothing more than a percentage of that ideal amount of fuel for perfect cylinder filling. So referencing the earlier Required Fuel value that we calculated as 17.6 (in ms), a value in the VE table of 50 would mean that we're asking for 8.8ms of fuel at that particular operating point.
Now with that all cleared up, we ready to address the difference between speed-density and Alpha-N. Fortunately it's pretty straight forward: speed-density uses manifold absolute pressure (or MAP) as the input for engine load, and Alpha-N uses the throttle position for load instead. In my humble opinion, speed-density is the superior method. Alpha-N works OK if you don't have a good MAP signal (for instance in an ITB setup), but if you CAN get a good MAP signal, then that's better for two reasons. First, it automatically compensates for changes in elevation. Going for a nice drive up into the mountains? No problem, as the air gets less dense, this is automatically reflected in the MAP, and Megasquirt leans things out for you. It's like changing out jets while driving! Secondly, I like that MAP is a more direct measurement of engine load, where as throttle position is one step further removed. (Incidentally mass air flow, or MAF, is one step better still, as it directly measures the mass of the oxygen entering the intake, which is why all modern OEM applications use it).
Note: it's important to also use (and calibrate) an intake air temperature (IAT) sensor when using speed-density. The intake air changes density with temperature (aka there's less oxygen available at a given pressure with warmer air as opposed to cooler air), so the input from the IAT allows Megasquirt to compensate for this. Other than calibrating your sensor, I don't think there is any other setup needed for IAT correction, unless you wish to set up a non-linear correction curve (not sure when this is necessary though).
Anyway, without further ado, here is what my VE table looks like:
Remember yous might need to look a bit different, depending on your engine/pistons/cam/etc! I'll explain a few of the more important regions of the table now though:
Idle - I have my car set to idle at about 900 RPM and there it's at about 40 kPa. I've found a VE of about 44 to give the smoothest, tad-rich (AFR ~14.0) idle. Notice that I keep this area of the VE table pretty flat with a lot of 44s so that a little fluctuation in idle speed or MAP signal won't drastically affect the injector pulse widths.
Cruising - This is in the mid-MAP range (around 40-70kPa) from 2000-4000 RPM. Here the VE tables are only in the 50-60 range, not a whole lot higher than idle, compared to the 70 and up for most of the higher load and higher rev area. While cruising down the highway or a back road with only light throttle, you can really run much leaner than when you're accelerating, and thereby get much better fuel economy. More on this is the AFR section but for now just remember to not get too aggressive with the VEs in this area.
Peak power/torque - Just the opposite of cruising, here at high load you want to error on the richer side. At full throttle above idle and higher RPMs with even a modest amount of throttle applied, I'm pretty much have the VEs at least up to 75, and they climb from there up to above 100 in the peak power band. You may also notice that my MAP axis goes up to 110kPa: above atmospheric pressure. This is because the 318i manifold has magnificent resonance properties in the mid-range (due to its long intake runners) and thus delivers a healthy bit of supercharging in this sweet spot, so I needed to run the table up into that range to capture this. Megasquirt will extrapolate linearly out beyond the VE table, but I prefer to try to keep everything under my control as much as possible. Note that for a turbo application, your VE table MAP axis should run all the way up to your max boost pressure, and VEs should also be increased accordingly.
Overrun - This is the high vacuum area at the very bottom of the table (20-30kPa) when the throttle is closed for deceleration. Obviously you don't need much fuel here, so the VE numbers stay pretty low. I do raise them a tad at the low RPM just so the area around idle is smoother, though in truth I'm not sure if seeing as low as 20kPa at <1200 RPM is even realistic.
7.) EGO and AFR table
If you're an experienced carb tuner, it's entirely possible to develop a pretty good VE table with just narrow-band or no O2 feedback (my dad and I did this on his 911 Megasquirt build) and achieve pretty good results. But in this day and age, using wide-band O2 feedback makes this infinitely simpler, and utilizing the AFR table and EGO closed-loop feedback in Megasquirt makes it so easy it almost feels like cheating.
NOTE: I STRONGLY recommend leaving closed-loop EGO correction OFF or at least restricted with very limited authority while you develop your VE table until it's quite good on it's own. Just last weekend on my way to the Vintage my EGR sensor started to fail and reading very (untruthfully) rich, which caused the EGO control to reduce fuel and made the car run LEAN! Fortunately the EGO only had 5% authority, so I was only running 5% lean. But if I had for example had the EGO dialed up to 30%, then it would have instead run 30% and that would have been VERY bad! You want a good solid VE table as you basis, with EGO just for tuning feedback and at the most fine adjustments while running.
The AFR table is just what it sounds like. It looks just like the VE table only here the numbers are your desired AFR target at the given operating points. Just like I mentioned for the VE table above, I like my idle area to be just on the rich side of stoichiometric, lean while cruising, rich under heavy load, and a bit lean on overrun, and smooth transitions between these areas:
Again you should decide what works best for your own specific car, but this should give you and idea for a starting point. Next, here is what my EGO control settings look like. With a well developed VE table, I allow it the authority to adjust the fueling amount by 10% to try and hit the AFR targets. While dialing in an early VE table, I would suggest starting with no more than 5% (have some EGO is helpful during tuning, as it allows you to see in the datalogs when the EGO is kicking in and how much). I also cut the EGO control out at idle (only active above 1200 RPM) and at full throttle (only active below 75% TPS or 90 kPa). This is generally considered good practice so that the controller doesn't end up 'chasing its tail' in a feedback loop in these sensitive areas:
8.) Spark advance table
Assuming that you're using Megasquirt to control your ignition (or even if you using something different like the 123Ignition distributor), you'll need to program in your desired ignition curve. The factory advance curve is a pretty good but conservative starting point, with about 15 degrees of advance at idle and 30 degrees of 'all-in' advance by 3000 RPM, and then an additional of ~5 degrees or so added for the overrun/high vacuum areas. My advance curve is slightly more aggressive than stock but still on the conservative side. I found the biggest improvement to come from a steeper climb in advance just above idle helps a lot with pulling off the line. I've think that 2002s tend to really like more advance in general, and I think with good 91 octane gas and a well dialed-in VE tune you can get all the way up to 40 degrees all-in, but play it safe starting out and increase slowly to avoid issues with pinging/detonation! I also found it helpful to increase advance just a bit *below* idle speed, so that idle sits in a bit of a 'valley' in the advance curve, which helps keep it stable:
9.) Datalogging and tuning
Congratulations, you now have all of basics setup and should be ready to try starting your car! Here's my recommendation of steps to follow for the first startup:
1.) Cranking - Double check your timing, cranking/ASE, and idle control settings and then make sure the engine at least catches after a small amount of cranking. It's OK if it doesn't want to idle yet, but if it doesn't want to fire at all then go back to the wiring and these three settings first.
2.) Idle - First aim for a faster than normal idle, maybe ~1200 RPM, and adjust the warmup and fast idle parameters to try and keep it in this range until the car is fully warmed up. Once it's warm, make sure all warmup enrichment and fast idle adjustments have all cut out, and then work on the VE table, timing, and the mechanical idle stop to get a smooth idle at your desired idle RPM with an AFR in the low 14s. After you have a good warm idle, you'll probably need to go back later after another cold start and re-tweak the warmup and fast idle settings.
3.) Accel enrichment - Once you've got the car warmed up and idling nicely, start playing with the acceleration enrichment settings to get good throttle response so that the engine revs up quickly when you blip the gas and then returns back to a good idle without drama.
4.) Datalogging - Great, now time for a test drive! But before you pull out of the driveway, start recording a datalog (built-in capability with TunerStudio). This will let you review everything that happens during your drive and decide where and what adjustments need to be made accordingly. Here's an example from one of my logs. Let's just focus on the top most of the three graphs. Here, white is RPM, red is MAP, and green is AFR (you may want to enlarge for better viewing):
In the first (left) half of the graph, I'm cruising and then decelerating. MAP is generally low, AFR is staying on the high side, and RPM is coming down with a few upward blips where I downshift. As soon as I come to a stop (low point on white RPM line), I then do a moderate acceleration in 1st and then close to full throttle for 2nd and 3rd (middle portion of the graph). If you look closely you can see a brief point early in the 1st gear run-up where the RPM drops just a little (I'm pretty sure this is due to the clutch engaging) and the AFR goes a bit high/lean - This is an area where I should consider increasing the value in the VE table! The gear shifts are the 'spikes' in the RPM line. Right where the indicator line is at I have lots of throttle (93.5 kPa) in 2nd gear at 4139 RPM and AFR is 13.2. In the tiny gauges at the very bottom of the screen, you can see that my target AFR (AFRtrgt1) in this area is set to 12.5, so I might want to richen this are up slightly, but 13.2 is already pretty good. The thing that does jump out at me in this section though are the low (rich) spikes in the AFR line right when I close the throttle (MAP goes low) to shift. This means I ought to decrease the amount of fuel during deceleration in my acceleration enrichment parameters (decel cut is just the opposite of accel enrichment for when the throttle is actively closing). In the last (right) section of the datalog, I shift into 4th and am back up to speed and cruising again and the AFRs nicely go back up the leaner cruising numbers. It takes some practice to get the hang of looking at and correctly interpreting all of this stuff in the datalogs, but THIS IS HOW YOU TUNE, FOLKS, so start grabbing some data and practicing! You should be able to get things running pretty well with just a few goes, but it will take many, many different rounds of test drives and the resulting finer and finer tweaks to get things close to perfect, but hey, this is part of the fun of DIY Megasquirt after all!
Good luck and I encourage you to post additional questions and/or datalogs in the comments if you need further help with any specifics!
On the way to the SoCal Vintage BMW Meet the new to me car was overheating when reaching highway speeds. The drive was a bit painful with multiple stops to let the car cool and took forever since i had to keep the car at about 80km/h. We didn't get to Santa Barbara until after 1am. Early morning and slow drive to Van Nuys followed.
The plan for the show was to set up the tents, put the cars in place, and then find a suitable radiator.
By the time we made it to Woodley Park, Ray from Elefant Supply already had the canopies set up. We did a little rearrangement, unloaded 3 tourings and had the FAQ booth functioning. Shirts, cars and tiny cars were on display, so it was time to find a radiator. Lucky for us Jack Fahuna had 3 at his shop, ProBimmer, about 5 minutes away. We were able to get the newly recored radiator, thermostat and a switch, just in case.
After some socializing and looking at cars it was time to finally get the radiator in place. The best way to get this done quickly is to have Jan and Alex Echeverri standing by.
Here are the photos of what happened between 1:00PM and 1:27PM
Removing clamps and disconnecting the hoses:
A well deserved drink
Removing the old radiator
Positioning the new radiator
4 bolts going in
Connecting the hoses
Installing the sensor
Running the engine to get the water flowing. Heater is on.
Using this Google spreadsheet developed by Grice Mulligan, AKA williamggruff, users and abusers alike can investigate the outcomes of matching various BMW transmissions and differentials, and calculate a variety of RPM and speed values that result.
Inputs are simple: Tire circumference and Redline RPM. Tire circumference can be estimated using the link provided. Redline RPM can be a target RPM, such as a preferred RPM at a cruising speed.
BMW transmission gear ratios and differential final gear ratios.xls
There are basically two different types of mechanical fuel pumps that were used on '02s. The '02 fuel pumps have 8mm inlet with 6mm outlet hose fittings, and one that has 8mm in and out hose fittings that is a performance upgrade......
#13 31 1 252 506 Pump is defined by its lever type operating mechanism that is below the mounting surface and the 8mm inlet and 6mm outlet hose fittings. This pump was used on single barrel and 2002 TI's before mid '72.
#13 31 1 265 192 The operating mechanism has a pin and coil spring that stick out of the mounting surface. This pump was used on 2 barrel '02s from mid '72 on.
#13 31 1 260 677 was used on the carbureted 6 cylinder M30 motors and is defined by the lever type operating mechanism, its ability to clear most side draft intake manifolds, AND its 8mm inlet AND outlet hose fittings, thus it has a greater volume capacity while having the perfect pressure for running many different types of carburetors on the M10 4cyl engines. M30 6cyl engines that used this pump had 170 to 190hp and dual carburetors. This is THE mechanical pump to use with side drafts on modified performance engines!
Picture of the two different operating mechanisms.
There are two different fuel pump push rods for 2002's. Early longer ones are for the "506" early pump and the "677" M30 engine types. This is the push rod for all the BMW mechanical fuel pumps that have the lever type operating mechanism that is below the mounting flange.
The two part #'s for the pushrods are...
#13 31 1 250 398 is 107mm long. Most of these have a narrow section in the middle of it. This is the push rod for all the BMW mechanical fuel pumps that have the lever type operating mechanism that is below the mounting flange. These longer pushrods are for the "506" early pump and the "677" M30 types.
#13 31 1 255 727 is 88mm long. This shorter pushrod is for the "192" and all the pumps which have the pin and coil spring mechanisms that stick out of the pump mounting flange.
#11 11 1 721 614 is the insulator block With gaskets on it.
#113 127 311 is a VW number for the gasket to reseal an old insulator block, you'll need two of them.
#13 31 0 075 752 is for the insulator bushings, use two (2).
To mount the pumps use two of the above bushings, two wave lock (not split lock!) washers and two 8x1.25 x 12mm wrench size nuts with a dab of blue loctite (13mm nuts don't work as nice in the tight space around the pump). Some try and use nyloc nuts but I find most of the time that the studs are not long enough to grab properly into the nylon. Try and not over torque the nuts as the bushings are fragile.
If your cylinder head doesn't have the provision for a mechanical fuel pump I suggest this Pierburg unit part# 7.21440.51.0 . They are specifically rated for the fuel pressure that our carburetors require, thus do not need an external fuel pressure regulator. And they are very quiet, especially when compared to a Facet pump. (SSF sells these Pierburg pumps as replacements for early carbureted Porsche 911's)
In addition to this electric fuel pump I suggest wiring it with an RPM sensing relay like that from an early E21 320i fuel pump relay part# 12 63 1 277 245, or this VW relay that has a built in fuse holder (15A) part# 321 906 059C . All these relays need to work is a tach signal from the coil. Even Tii's can be wired with one of these relays to make them safer.
Mike's 318i Electronic Fuel Injection setup including his custom cold-air box
This conversion is not recommended for anyone who is not mechanically knowledgeable and ready for a challenge. There are many steps involved and several places where fabrication skills will be required. If you feel that you are not ready for this, it would be best to leave the conversion to a professional.
When considering this conversion, it is advisable to do significant research to familiarize yourself with the theory, parts and workings of both your 2002 and the 318i fuel injection system. There are several valuable resources available, namely ‘Bosch Fuel Injection & Engine Management’ by Charles O. Probst, SAE, ‘BMW 3-Series Service Manual’ published by Robert Bentley, and ‘BMW Coupes and Sedans, 1970-88 Repair Manual’ by Chilton’s. Other sources of valuable information may be obtained by locating the Electronic Diagnostic manuals for the 1984-85 BMW 318i. I found access to them from a local repair facility.
The more you know about all of the systems involved when doing this conversion, the better. Once can never have too much knowledge when it comes to tackling this type of project.
There are many systems involved in the conversion and many of them interact with one another. The systems involved are the fuel system, electrical system, vacuum system, ignition system, starting system, intake system, etc.
Below is a list of the components required to complete this conversion. This is only a guide. Some conversions will require more parts, others will require less, depending on your decisions regarding some critical questions.
From a 1984-85 BMW 318i:
Intake manifold w/ cold start sensor
Throttle body with throttle switch
Water inlet with sensors
Fuel injection rail with injectors and fuel pressure regulator
Complete engine harness with relays and ECU (make sure that the harness and ECU are from the same vehicle since they changed from 84 to early 85 and again in late 85)
Air flow meter
Main Fuel pump
Throttle connector for cable
As many hoses as you can get and don’t forget the hose clamps!
Intake air boot
Rear starter bracket from a tii, 318i or 320i or modify yours
From a late 320i:
In tank fuel pump/level sender and harness (measure your tank and get the correct length pump) I used 16 14 1 179 423 on an early ’72.
Aux air valve (0-280-140-126)
Throttle cable (custom made or from Honda)
Gasket for throttle body
8mm fuel injection fuel hose (the expensive stuff)
3/8 fuel hose
5/16 fuel hose
4 bay fuse holder
Fuses for fuse holder (7.5A, 15A & 20A)
Plug for water port in engine block
Blank off plate for fuel pump on head
Miscellaneous terminals and wire
Miscellaneous hose clamps
Battery relocation parts:
Battery cable (highly suggest getting one from an E30 with trunk mounted battery), battery mount, battery Air cleaner (cone style or stock) ECU mounting standoffs/bolts/screws Manual idle valve Miscellaneous hoses for pluming idle valve and aux air valve
Optional parts: (1984-85 BMW 318i):
Idle control module
O2 sensor (heated) (11-78-1-715-263)
O2 sensor mounting bung
Idle control motor
318i Distributor & camshaft
Ignition control module
Head gasket set
002 or 008 distributor
Metal fuel return line from mid’72 and newer 2002
There are some questions that you need to answer prior to delving into the conversion. This is by no means all of them either, there may well be many more, but this is a good start.
Are you going to use all the stock 318i systems? Such as idle air circuit, fuel pump relay, ignition (including distributor and ignition control module), O2 sensor, etc.
Will you be using your stock distributor or a 002/008?
See the list of optional parts above and add or subtract parts as required.
The first step in the conversion is to make room for the EFI fuel meter and other parts of the system by moving the battery from the engine compartment to the trunk (or under the back seat, if you get a sealed battery). This is easily done with a kit available from any number of aftermarket suppliers, welding supply stores, or you can use a cable from a junked e30. Check out the FAQ article on this procedure.
I suggest using the cable from an E30 with the battery in the trunk. This will allow you to have all the power sources you will need already available in the engine compartment without having to overload the starter terminal or the alternator. Approximate time to complete this step: 6 hours
Starting with the basics, you need a way to get highly pressurized fuel from the gas tank to the injectors on the engine. In my case, I had to install a steel fuel line (normally the fuel return line on newer cars) from a newer 2002 (later than mid-’72). I used cushion clamps available from a hardware store along with large-head aluminum pop rivets and silicone to install the fuel line. It is very easy to install, but takes some time lying on your back unless you are fortunate enough to have a lift. Approximate time to complete this step: 2 hours, or zero if your car already has the steel return line that you can convert to a feed line.
The next step is to install the fuel pump/s. There are a couple of ways to do this, but probably the best is to simply use the single in-tank pump from a 1991 318i. This is a high-pressure feed pump that eliminates the need to use an in-tank pre-pump and the seperate external high-pressure pump.
If you use seperate lo-pressure in-tank, and an external high-pressure external pump (which is what I did), the external pump must be mounted so it can be easily installed and be relatively out of the way of things like exhaust and suspension components. I installed mine using an old coil mounting bracket with a couple of rubber air cleaner mounting studs that are rubber mounted. (13-71-1-272-495) This way, it is out of the way, up near the differential, allows easy access and is relatively in line with the metal fuel line. Approximate time to complete this step: 2 hours with wiring, which we will get to in a bit.
Installing the in-tank pump and sender assembly is relatively straightforward. It is installed in the same fashion as the stock '02 sender with the exception that it is 2 pieces. If you have the one-piece 320 version, throw it away and get a 2 piece one. The one-piece version was replaced with the 2-piece version a few years after the 320 was first produced. Also remember to get the harness connectors that go with the pump and sender. The pump is a 2 prong and the sender is 3 prong. They are special connectors and are hard to find by themselves. Approximate time to complete this step: 1.5 hours with wiring
Next install the fuel filter. I got a stock 318i fuel filter and mounted it using a hose clamp. I drilled holes in a piece of strip aluminum and using pop rivets, fixed the hose clamp to the bottom of the car, near the fuel pump. Then I just tightened the clamp around the filter and installed fuel line between the pump and filter and from the filter to the metal fuel line. Remember to use high-pressure fuel injection hose because the fuel is under ~45psi at this point. Approximate time to complete this step: 1 hour
Once the in-tank low-pressure pump is installed, you can run your car with it (if it is still carbureted), just hook up the stock fuel hose to the outlet, wire up the pump and away you go. I ran my car with the stock engine-mounted pump until later in the conversion, but the fuel was drawn through the in-tank pump without any problem. Remember that there are two fuel connections to this pump, a feed and a return.
Fuel System Wiring:
Wiring the fuel pump/s can be a little tricky. If you decide to run the fuel pump relay, it would be wise to wait until later in the conversion to make the final wiring hookups in the engine compartment.
The connector for the in-tank pump has two terminals in it. One is green/purple and the other is brown. Brown is ground and the green/purple is power (+12V) The in-tank pump requires a 7.5A fuse.
The main pump has two connectors on it. One is power and the other is ground. If it was removed with the wires attached, power is green/purple and ground is brown. This pump requires a 15A fuse. The stock 318 system used one wire for both pumps and used a 15A fuse for both, but I wired them independently and used separate fuses for each. The fuel level sender has 3 terminals. One is signal, one ground and the other is low fuel. The brown wire is ground, one of the brown/gray wires is for level and the other is for low fuel. You may have to do some testing with a VOM meter to be sure you hook up to the correct wires. I installed a low fuel warning LED at the lower end of the fuel gauge. It works well and tells you when you are within 1 gallon of being empty. It is a grounding circuit, so hook power to your light and use the connection on the sender assembly for the ground.
Next was to install the 4-position fuse block under the hood. I chose a location close to the original fuse block on the firewall. The 4 fuses are for: one each for the fuel pumps (7.5A & 15A) one for +12V key switched (15A) and one for +12V key switch/crank (20A). Approximate time to complete this step: 1 hour
ECU and Engine Wiring Harness:
I installed the ECU in the glove compartment using standoffs from the bottom of the box to the ECU. I mounted it so that it moved with the glove box, since it would take more effort to install above the box similar to the 318i and there was limited space. Approximate time to complete this step: 2 hours
The next large step involves the main wiring harness. This harness is large and bulky. If there are systems that you will not be using, I would suggest that you remove all unnecessary wires from it prior to installing it in your car. It will be necessary to drill a large hole through the firewall on the passenger side of the car for the connection to the ECU. Retain the stock rubber grommet and use it to seal the hole. The hole is approximately 2” in diameter and can be made easily with a hole saw.
I would recommend that you acquire a complete wiring diagram for the specific year 318i you are using. The 318i changed from 1984 to early ‘85 and again in late ‘85, so finding the correct diagram is critical for the harness that you have. If you are going to use an O2 sensor and are not using a stock 318i exhaust manifold, it is recommended to use the late 1985 harness and ECU. This will allow you to remove all unnecessary wiring and be able to connect to the correct wires for power.
Click here for a larger version of this great wiring diagram provided by Ed Weimar. Ed got Mike's help on his own 318i injection conversion and created this diagram. Be aware that wiring colors do vary between the '84 and '85 EFI systems, so you should still get the appropriate-year electrical manual for the injection system you are using.
Wiring the harness into the 2002 is the hardest step in the process. I recommended that you check and recheck the wiring carefully. I removed the ignition wiring (ignition control module and vacuum advance solenoid), idle speed control unit (with the idle air motor), but retained the stock fuel pump relay. If you plan to use the ignition system, I would highly recommend you change the cam to use the 318i unit. It will give you the best control of ignition short of replacing the whole thing with and aftermarket adjustable version. Short of that, use a 002 or 008 distributor from a tii. It has a curve close to what the 318i has and will be the best stock approximation.
Once the harness is stripped of all unnecessary wiring, it can be installed in the engine compartment. I welded the mount for the fuel pump relay on the inner drivers’ side fender similar to where the 318i had its mounted. It worked well for locating the harness. Approximate time to complete this step: 4 hours
You can now hook up the wiring from the harness to the 02. You will need two power sources, an ignition source (- side of the coil), a crank source (starter) and ground. Basically that is all. Although it sounds simple, it takes some time to find the correct wires and hook them up. Approximate time to complete this step: 2+ hours
I used an O2 sensor in my conversion and installed it in the lower "Y" of the stock exhaust collector. Initially I used a non-heated sensor and was getting poor results since it was not getting hot enough. I changed to a heated O2 sensor and the system worked great. If you use a stock 318i exhaust manifold and down-pipe, you can use a non-heated O2 sensor. Either way, I would recommend a heated O2 so you can be sure the signal is accurate. Approximate time to complete this step: 1 hour
The throttle cable that I used was a custom cable made by Ireland Engineering in Southern California. I sent the stock 318i parts (at the throttle end) to them and had them fabricate a cable using those parts so it would fit the stock bracket. I had to fabricate an L bracket that I mounted to the brake booster mounting to hold the lower end of the cable housing above the pedal linkage. A simple rod end with a through bolt was used to attach the cable to the pedal linkage. I have also heard that a Honda cable will work, but do not know from what vehicle it is. Approximate time to complete this step: 1.5 hours
When I received my throttle body, it was not in great condition. I rebuilt it and made it work as new with only a two-hour investment of my time. I also rigged up a pump and tank assembly to test the fuel injectors. Since they were not mounted in the intake manifold, this was a simple arrangement of fuel source (gas jug), high-pressure pump and a battery. A simple momentary switch and 12V were required to power the injectors one by one. I verified that all the injectors were spraying a good pattern and that they all work. That was one thing that I did not want to be a stopping point in my conversion. If any of your components were from an unknown source, this is highly recommended.
I then installed the injectors in the intake manifold using new o-ring seals (13-64-1-286-708), two per injector. Approximate time to complete this step: 1.5 hours
As we approach completion, the intake manifold swap is next. It is a straightforward swap. Remember to use the new 318i water inlet. You will also need the 318i water pipe that runs under the intake manifold for the heater hookup. On a stock 318i there is a rod that supports the intake manifold above the starter. I was not able to make this fit even with a lot of bending and modification work. In the end, I left it off. It has been there since late ’99 without a problem. If it has not been done, it will also be necessary to use a starter bracket from a tii, a 318 or 320 which is different from the carbureted 2002. The extra tab for the manifold will get in the way of the throttle body. Approximate time to complete this step: 6 hours
Fabricate a bracket to mount the air flow meter to the drivers’ side fender well or the engine. I was able to make a simple bracket by bending some thin sheet metal and welding in a cross brace. I bolted it to the fender. This turned out to be a good move since it sits above the oil filter and needs to be removed to gain access to the oil filter. The installation of a K & N cone style filter completes the installation. I removed some of the radiator support to allow a direct line to cool outside air. Approximate time to complete this step: 2 hours
I utilized an auxiliary air valve and a manual idle valve in place of the stock idle air motor and controller. This is a very simplified way of removing the expensive electronic components (that are known failure items) and retaining the drivability and cold starting which makes the conversion desirable. It is, however, quite the plumbing nightmare. It took several tries and different configurations to make the two parts work together. It was not too difficult with the intake off the car, but it was a little cumbersome once it was installed on the engine. It does, however, work very well and is less expensive to replace than the stock 318i components. Approximate time to complete this step: 3 hours
Final steps are to hook up all the wiring and hoses to the new components. Once all the hookups are made, check over everything carefully. Now you are ready to take your fuel injected 2002 for a test drive assuming you have done everything correctly.
After spending nearly 55 hours on the actual conversion and countless others on the internet, reading through manuals and books, searching for information from other sources, talking to numerous people and deciphering all the information, I am very pleased with the outcome of the conversion. It makes driving my 2002 daily a pleasure (not that driving one is anything but).
With the reliability and performance that I now have, I believe that I can drive this car for many years with few problems. There are also other things that can be done to boost the performance even more, namely turbo charging, but I will leave that to someone else. Other ideas are to use other throttle bodies, convert to LH-Jetronic, which uses a mass air sensor instead of the air flow meter used by the L-Jetronic. All of these and more are possible modifications of the existing system. I have not done any of them yet, but who knows, maybe in the future.
Once I had gotten the basic system working well, I was initially impressed with the performance of the FI system. However I had heard pinging numerous times, specially at high RPM. I adjusted the timing (stock '02 distributor) until the pinging ceased, but noticed poor idling. This led to the installation of an Electromotive HPV-1 ignition system that is adjustable.
In hindsight, it would have been better to install the Electromotive TEC-II or SDS system which controls both ignition and injection or to have installed the stock 318i camshaft and distributor.
Both of these solutions would have provided better control of all the systems together, although both the Electromotive and SDS systems would be far superior in that they are fully adjustable.
I start by saying always liked "stock look” under the hood and never cared for appearance of Weber 32/36 air filter on my 76 ’02. On the technical side, Weber set up is such that air drawn into intake manifold is always warm/hot (from engine compartment surroundings). So it is opposite of air induction system was originally designed for base model. On original setup for most parts selection of warm and/ or cold air determined by air regulator housing. The original system always let outside ambient temperature air to intake, which it is believed to make engine run smoother and better fuel efficient.
Stock air cleaner and weber carburetor Integration has been discussed on this forum previously and there are lots of good information how to do this, but did not find much information on purpose of air regulator box, heat shield over exhaust manifold. So I had to learn about each component as I was acquiring them and at the end found system design quit simple and efficient. So it was learning experience and then I thought why not write article about it.
Procedure is written for a 1976 model year car, which I think will be very similar for earlier model years as well, but don’t know that for sure. As I mentioned before I did not have any of original components so I had to place add on this forum under “Part Wanted” section and thanks to members that helped for me to acquire needed parts.
After several search and reading up on this topic it appeared that there are at least two methods to accomplish this job if not more; one is to use an adaptor between air filter housing and carburetor; Second option is to not use an adaptor. I don’t know pros and cons of one system to another, but as I recall If none adapter version is decided then a fabricated closed-cell foam gasket should be installed between air cleaner and carburetor. Below pictures (not mine) for none adaptor version for reference.
In my case I opted to go with an adaptor version.
Here are photos of components that were needed;
JAM air cleaner adapter Picture (13/36)
Fabrication and Installation:
Turn-over air filter housing, remove rubber seal and slide the adapter over the air cleaner flange
Trace JAM adapter inside and three mounting holes onto air cleaner
Remove adapter then remove excess material by various tools (i.e. tinsnip pliers, dremel tool)
Drill previously marked holes with ¼” drill bit
May want to prime/ paint modified area to prevent future rust
Disconnect cable from battery negative post as a precaution
Remove Weber air filter assembly including filter base plate, four screws and breather hose
Fasten two brackets to intake manifold (picture 13/39, item #1 & 8)
Slide down new gasket ontop of carburator
Set JAM adapter over carburetor and then place gasket on top the adapter
With air cleaner over JAM adapter line up 3 mounting holes.
Loosely fasten air cleaner and JAM adapter to carburetor by using M6x45 screws and flat washers supplied in JAM adapter kit. Do not tighten screw yet.
Fasten air cleaner to the brackets that were previously fasten to intake manifold. Purpose of these struts are to partially support weight of air cleaner and not all air cleaner weight on carburetor alone
Using piece of none braided hose connect cylinder head valve cover vent opening to air cleaner tube opening
Remove air regulator box cover screw and withdraw the cover. Oil the valve pivot points and check adjustment.
Replace cover and screw then install the box onto snorkel (right rear of the radiator)
use rubber boot to connect air cleaner to air regulator box
Set new filter in and put the cover on and close clamps
Slide heat shield cover over exhaust manifold and fasten it with two M6x16 bolts along with flat and wave washers. May want to apply anti-seized compound to the bolts for ease of removal just in case later on
Connect heat shield to air regulator box. Curved tube (picture 13/36, item#6)
Slowly try to close hood and absorb to see if hood inside makes contact with top of air cleaner housing. I did not have any interference issue, but if it seems to be the case, you may want to consider followings for remedy;
Have machine shop to shorten JAM adapter height by not more than 2mm
Thickness of isolation gasket sandwiched between carburetor bottom and intake manifold opening
Foam pad inside bonnet
Raising bonnet slightly
[*]Re-connect cable to battery negative post
[*]Have someone to assist with starting the car and ensure hood and air cleaner top not making contact as revving up engine, which in this case engine wants to tilt to right hand side
[*]Last but not least need to paint air cleaner so it “looks stock” but have to wait till weather warms up around here.
Air Regulator Housing Adjustment
With the valve lever in “Summer Position” (vertical) air is drawn only from outside the car
Release lever from “summer position” for other seasons.
With the valve lever in the horizontal position, air drawn in from front of the car is mixed with air heated by exhaust manifold in specific proportions by the action of bi-metallic element which is dependent upon the current engine and ambient temperatures.
Original Author: Steve Kupper
If your shifter is a bit looser then you'd like, and you want it to feel like new again (do you know how it felt when it was new?) here are a few simple and very inexpensive things you can do to get the like-new effect. This article is based on a late 2002 style shifter. Please look at what yours looks like before ordering parts.
Few more parts needed (don't have them yet): 8 Spacer ring 25 11 1 220 199 9 Circlip 25 11 1 220 379 10 O-Ring 25 11 1 221 243 11 Spacer Ring 25 11 1 220 439 12 Circlip 25 11 1 220 379 13 Tension Bush 25 11 1 203 682 14 Rubber Washer 23 41 1 466 118 15 Dowel Pin 23 41 1 466 134 16 Gear Selector Rod Joint 25 11 1 220 198
UUC Motorworks Short Shift kit. (Kit will include Spacer Ring 8 and Circlip 9.)
UUC Motorworks Cartridge Bearings
Other Short Shift kits are available, but this is the one I happened to have used and liked.
Time needed: about an hour
Lift the car and put it on jack stands. The usual disclaimers apply. Use a solid, flat, ideally level surface and jack from under the front subframe with a block of wood on top of the cradle to protect the subframe. Place the jackstands under the stamped steel frame rails that are welded to the front floors, again with a short piece of 2x4 wood on the stands to protect the frame rails and spread the load.
Once the car is SECURELY in the air, get under it and remove the Circlip from the shifter. (It might be a good idea to print out the diagram included in this article so you know where everything is.) Slide the Shifter out from the Selector Rod. Remove the Gear Selector Rod Joint by removing the tension Bushing and then sliding out the Dowel Pin.
Now go inside the car and remove the shifter boot. Remove the Circlip and the Spring and then pull on the Shifter. It might require a bit of force, but be careful not to hit yourself in the teeth.
Go to the workbench and remove the Selector Rod from the Gear Selector Rod joint by removing the Circlip.
Connect the Gear Selector Rod to the Joint using the O-Ring, Spacer Ring and Circlip. Wet the Rubber or Sponge Washer with oil and slide it into the Joint.
Get under the car and connect the rod to the transmission using the Dowell Pin and Tension Bushing. This is the hardest part of the whole installation. It might take you a while to aim the Dowel Pin correctly. And it is hard to do with the driveshaft in place.
Get into the car and install the plastic bushing and the Shifter.
Get under the car and slide the Selector Rod into the Shifter and attach it with another Circlip. Make sure to remember to use a spacer ring.
Pull the Shifter boot back in place and you are all done. Enjoy the like-new shifter!
Delrin Bushing Installation:
It may be necessary to remove part of your exhaust system to have easier access to the bottom of the shifter. Cars are set up differently and abilities of individuals differ as well. It should not be necessary to remove the shifter from inside the car.
From under the car, remove the allen socket set screw with the appropriate allen wrench from the rear of the shifter shaft. Remove the cross pivot pin, the transmission shifter yoke will slip down yet remain attached to the transmission. Remove the bushings in either side of the yoke by pushing them in one at a time. Replace with the new bushings and install reverse of removal.
FITMENT: These bushings are made slightly thicker on the bearing surfaces (colored purple) and these surfaces may have to be sanded down with common sandpaper to ensure a loose enough fit to place the shifter shaft into the transmission shifter yoke when the bushings are installed. Trial fit your new bushings, if they are too tight, sand the purple surfaces a little at a time and refit. This will give the tightest most precision shifting action in YOUR BMW 1600/2002.
If you have any questions, post them on the Message Board!
Some basic information about the Solex 36 - 40 PDSI or PDSIT. Jets, parts and tuning. Most of the parts are no longer available, but the rebuild kits can be found all over ebay plus many 2002 parts vendors.
Solex 36 - 40 PDSI and PDSIT were available on the early 1600, 1800 and 2002 models. PDSIT is an automatic (water) choke version and is hard to find. The venturi and jet differences are listed bellow.
read all my Neue Klasse pages in my photobucket - some maybe
Spray castle nuts with penetrating oil several times over a course of a few days and let it to break loose surface tension rust and corrosion.
Jack up and support vehicle
Remove cotter pin
Replace tire and lower car to ground
Apply parking brake and car in Neutral
Loosen up castle nut with 36mm socket and long handlebar
Jack up and support vehicle again and release parking brake
Remove tire, castle nut and drum brake
Heat hub for about 15 minutes. Rotate hub as your heating it.
Pull off driving flange/hub with extractor
Detach output shaft / CV shaft from axle shaft and tie it up
Screw on castle nut with notches facing brakes, then use soft hammer to drive axle shaft out
Pry out inboard and outboard sealing rings. From inboard side drive out outboard bearing with soft punch (brass) and hammer. Now shim ring (if any) and spacer sleeve can be removed from outboard side. Now can easily remove inboard ball bearing with punch and hammer. It is imperative that circular shim and spacer sleeve for each wheel kept separate if rear wheels bearings removal done at once.
Wipe inside with paper towels and spray inside with brake cleaner.
New bearings and seals
Next, Pack bearings with grease and grease sealing rings lip
Install inboard bearing. I used 1-1/2” dia. PVC coupling and plug to drive bearing in against bearing bore stop.
Coat sleeve spacer outside with 35 grams grease and then insert sleeve to the cavity from outboard.
Insert shim ring then, install outboard bearing making sure it is seating against shim
Install inboard and outboard seals
Insert axle shaft from inboard side all the way in
Install hub/ driving flange. May need to use hammer on face of hub to seat it in
Tighten castle nut with 36mm socket
Replace drum brake
Install tire and lug nuts
Put rear wheels on ground
Apply hand brake
Secure front and back of rear tire with objects to prevent it from rotational movement
Tighten castle nut to specified torque
And for last time Jack up and support vehicle
Install cotter pin and bend tabs
Replace output shaft / CV shaft to axle shaft flange
Put tire back on and tighten lug nuts
Lower the car and torque lug nuts to 65 ft-lb
Rev. A: attach CV shaft to axle shaft flange (08/13/2015)
I've seen several mentions of plans to use the 320i inner cv joints when installing a limited slip diff from a 320is but didn't find any posts about how it went. Most were aiming to avoid both using spacers and redrilling the diff output flanges. I was only concerned about avoiding the latter. I decided to take a few photos along the way to document what I found. There are lots of articles about other aspects of the swap so I'm just focusing on the dimensions of the axles below to supplement what's already out there.
Redrilled output shafts would be a much easier approach with minimal downtime if your 02 CV joints are still good. I wasn't so lucky.
I knew I had one bad '02 CV joint and the other three had dry, cracked boots so I'm repacking four cv joints regardless.I purchased the 320 halfshaft assemblies with the differential. The 320i joints looked pretty good, without any tears in the boots but they were cracked and due for replacement as well. I figured I could salvage two good 320i joints, and at least two of my '02 joints should be good.
Lots of posts I read on the topic were asking if this approach would allow installation without spacers. I didn't find any posts confirming it would work so I ordered a pair of IE spacers drilled to 10mm for the 320 cv bolts, and a longer set of bolts just in case.
Donor 320is for the axles and diff was an 83. My 02 is a 74. Earlier years of 320i used thicker CV joints that would require a different solution;
33 21 1 205 746 ..... 2002 & 320i >08/1978 (M8 bolts)
33 21 1 207 277 ..... 320i 09/1978>09/1979 (identical to above except M10 bolts)
33 21 1 208 439 ..... 320i 09/1979> (narrow width & M10 bolts)
1976 2002s are unique as well. You'll apparently have it easier than most by just swapping your 2002 diff flanges into the e21 diff.
Here are the axles side by side. All my photos are before or during disassembly and cleanup so please don't give me a hard time about how dirty everything is. :-)
The 2002 axle is on top, the 320i on the bottom. CV joints are fully compressed for consistency. As you can see, the 320i halfshafts are quite a bit shorter overall than the 2002's. The CV joints themselves are quite a bit thinner, too. I started to question whether this was going to work but I decided to press on and see what the axles look like.
Here are the bare axles;
The 320i axle is on the right. Quite a bit longer than the 2002 axle on the left. There's still hope!
Next, I test fit a 320i cv on one end, and an 02 cv on the other end of a 320i axle. My other '02 halfshaft is shown here for reference. Again, both are fully compressed.
They're the same length! Since the 320i differential is narrower than the '02 diff, this means I'll probably use the spacers.
I checked fitment on the car and while the CV flanges could reach the diff and stub axle flanges without the spacer, the spacer moved the cv joints closer to the center of their range of extension both at parallel to the ground (shortest flange to flange distance), and at full droop (greatest flange to flange distance).
10/15/15- Checking back after ~1 year and a few thousand miles. Everything is still working great.
Having recently sorted out three tiis that had sat for varying lengths of time, this is now the recipe I would advise for either sorting out a fuel delivery problem or resurrecting a long-dormant tii. I just did it with Old Blue, the '73tii I just bought that had been sitting for ten years.
Pull the pickup tube from the gas tank.
If, as soon as you open the tank up, it smells like varnish, you already know that you're going to need to systematically clean everything -- at a bare minimum, you'll need to drain the old gas and blow out all the lines.
Inspect the screen at the base of the pickup tube.
Verify that both the outflow and return lines on the pickup tube aren't clogged (I've just seen three tiis with this problem). Ream then out with a coathanger and compressed air.
Shine a flashlight in the gas tank and make sure it's not full of rust or sediment. I've seen them look like pot roast. If it's bad, pull the gas tank and clean it. At a minimum, pressure wash it, dry it, put it back in, and refill it with five gallons of clean gas.
Pull the fuel pump and inspect the conical screen at the inlet. It may be clogged or completely missing.
If the fuel filter inlet screen is missing, tap the inlet side out onto a paper towel. If rust and sediment come out, I'd recommend you replace the fuel pump.
Disconnect the fuel filter to the left side of the radiator.
With the fuel pump and fuel filter disconnected, blow the main fuel line out with compressed air into a bottle to catch what comes out. Inspect it. If there's massive amount of rust, blow brake cleaner into it and repeat until the rust is no longer visible when blown into a clean rag.
Undo the return line from the back of the Kugelfisher pump.
Do the same blowing out of the return line.
Remove the pressure valve from the back of the Kugelfisher pump and visually inspect it, looking through it against a bright light. There should be a pinhole of light visible. If there's not, clean it with brake cleaner until there is.
Pull the banjo bolt out of the front of the Kugelfisher pump and inspect the barrel-shaped screen inside it. I've spent hours removing and cleaning them.
As said above, if the fuel smells like varnish, you really should blow out the plastic injection lines with compressed air, and pull the injectors and have them cleaned and tested.
Reassemble everything, preferably replacing every fuel hose -- or at least every fuel hose that is too soft or rock-hard -- with OEM.
Put a fuel pressure gauge just before the Kugelfisher pump. Turn on the ignition to run the fuel pump. It should read 29psi.
Inspect every part of the fuel delivery system for leaks.
Try to start the car. Look in the throttle body at the cold start injector. If no fuel is being squirted, you'll have to troubleshoot the thermo time switch, or simply wire the cold start injector temporarily to the battery, or semi-permanently via a switch.
Start the car. Inspect the plastic injection lines carefully for leaks, both at the base of the lines at the Kugelfisher pump as well as in the lines themselves (they do crack with age).
Look in the throttle body at the cold start injector to make sure it's not leaking.
If the car still doesn't run right, follow the procedure in the "BMW 2002 Tii Injection Manual" to the letter to isolate the problem to the delivery valves, the suction valves, or the injectors. Add to that the information in my article about pulling the head off the Kugelfisher pump. Don't pull the head if you don't need do, but that article (and other posts) tell about removing the delivery and suction valves and verifying, using a small wooden dowel, that the plungers ("pistons") inside the Kugelfisher pump's head aren't stuck and are free to move up and down. But worry about the injectors first. Old varnished fuel MAY sit in the Kfish pump head, but it DEFINITELY will sit in the injectors.
I wanted to make this a separate thread from the long one about my fixing Brian Ach's rust-contaminated tii to make it easier for other folks to find the pertinent information without having to look through pages and pages.
Pursuant to the other thread, I was trying to nail down a rough running problem in a tii. It eventually turned out to be four rusty injectors, one of which was internally broken, but before I knew that, I convinced myself that rust in the fuel had created a "stuck piston" inside the Kugelfischer injection pump, without really even fully understanding what that meant. There are a few threads that describe the phenomenon, and one that says "it's not really that big of a deal to pull the head off." So I did.
I learned so much. Let me try to boil it down and serve it up.
Read the Tii Manual, But the Cutaway Photos are Misleading
If you have a tii and haven't read "The BMW 2002 tii Fuel Injection System" by BMW of North America (http://www.2002tii.org/pump/pump_guide_v1.pdf) cover to cover, just stop right now and do it. It's an excellent resource. However, it relies, perhaps overly, on those funny cutaway diagrams. The manual essentially says "there are no user serviceable parts inside the pump," so they don't say a word about unbolting the head. Because of this, the manual shows you no photographs whatsoever of the stand-alone head or the pump with the head off, and yet it shows you these odd cutaway views that are a little misleading because, obviously, when you open it up, it doesn't look like that.
So here's two pics of the pump with the head off, first with the pushrods exposed (this will be explained more below):
and then with the pushrods partially obscured:
And here's a pic of the top of just the bare head. The four holes down the middle are where the four suction valves and the Allen key caps on top of them normally go. The two sets of two offset holes outside these are where the delivery valves normally go. The injection lines screw onto the tops of the delivery valves.
and the underside, where you can clearly see the bores (cylinders) that the plungers (pistons) run in:
The Kugelfischer Pump Is and Is Not Like a Little Engine (engines don't have pistons IN THE HEAD!)
The manual says "The injection pump is Kugelfischer Model PL 0 Mini Pump. Its interior is very much like a small engine." Well, it is and it isn't, and we'll get into that.
The description in the manual of the basics of fuel delivery is quite concise:
"There is a camshaft which drives tappets. These, in turn, drive plungers which are pushed back down by springs. When these plungers move down, suction valves are pulled open and gas is sucked into the cylinders of the pump. Pressure created by the upward stroke of the plungers forces the suction valves closed and the delivery valves open. Fuel is pushed out into the delivery lines and through the injectors. The pressure generated by the injection pump is quite high -- 35 to 38 bar."
And that is really all that you need to know about fuel delivery, which is all I was concerned about. The rest of the action that goes on in the bottom part of the pump has to do with enriching, fuel timing, warmup, and other things.
But here's what it doesn't say, and what you only learn when you pull the head off:
--There's some truth to the analogy that the pump is "very much like a small engine" because the description talks about a "camshaft" and "tappets" and "cylinders" and "plungers" (note it doesn't say "pistons"), and it does have those pieces.
--However, the analogy falls flat because there isn't really a block, the "camshaft" and "tappets" are in the bottom part of the pump, and the cylinders are in the head, which is all topsy-turvy from the "little engine" image.
How Fuel Delivery Works (the head is pretty dumb, and they're spring-loaded plungers, not pistons)
--The manual refers to plungers, but many people refer to them as pistons. And since they run in the little cylinders, people think of them as pistons. But they're not rigidly connected to the camshaft. There aren't little connecting rods and wrist pins or anything. The little pushrods ("tappets") shown in those first two photos of the bottom end of the pump go up and down, and push on the bottoms of the spring-loaded plungers. The tops of the spring-loaded plungers go into the cylinders on the undersides of the head. The springs cause the plungers to retract from the underside of the head.
Here are the little spring-loaded plungers. Each is about half the length of my pinkie:
And here are the plungers sitting in the head (the whole assembly is, obviously, sitting upside down):
--Other than the plungers going up and down, the head is really quite dumb. It's really just passageways.
--The suction valves sit on top of the plungers under those little Allen key caps. You can undo the caps and pull the suction valves out with fine tweezers (NEVER USE A MAGNET) and see the tops of the plungers. You can then push down on the tops of the plungers with something non-marring and the plungers will just bounce on their springs. Again, they're not rigidly attached.
Here's a photo looking down into the holes where the suction valves sit. They've been removed, so you can clearly see the tops of the plungers. The delivery valves are out as well so you can look into their little threaded chambers, but unfortunately you can't see the passageways.
--Fuel enters the head at the banjo bolt at the front (not shown above; not installed) and leaves at the pressure regulating valve at the back (shown above on the back -- right -- side of the head). In between, it's just a straight shot. Except, as the description says, if a plunger is drawing downward, the suction valve allows fuel to be drawn into that cylinder.
--As the plunger goes upward, the fuel is put under pressure, and is squirted out a passageway in the side of the cylinder wall. That's right -- fuel comes in at the top but is squirted out the side. Rather unlike a little engine, huh? The passageway goes to the small threaded chamber that the delivery valve for that cylinder is screwed into.
--If you do what I did (and, seriously, don't do this) and remove one of the delivery valves and run the fuel pump just to make sure the passageway is clear, fuel will shoot quite spectacularly (clear across the garage, in fact) out of the passage in the threaded delivery valve chamber at about a 45 degree angle..
Why Rusty Fuel Can't Contaminate the Rest of the Pump
--Again, the bottom end is where the mechanical smarts are. The head is really kind of dumb, just passageways. Fuel comes in the front, some of is squirted out the delivery valves, but most of it flows out the back of the head at the pressure regulator valve.
--As per the other long thread about Brian's car, I can't comment on how rusty gas got into the injectors, because I don't know. I also can't say for sure whether, if rusty gas inundates the pump head, if it should be expected to pass through the suction valves, the passageways in the head, the delivery valves, and train-wreck at the injectors. But I can say that rust or other contaminants in gas really have no way I can see to get into the bottom Swiss watch part of the pump.
--And it is very straightforward to pull off the pump head and just blow the passageways out with carb cleaner and compressed air just to be certain they're clean (which is what I did when I had it off).
The Fallacy of the Stuck Plunger
--I honestly thought I had a stuck plunger because I followed the procedure in the manual to trouble-shoot rough running in one cylinder that seemed to be fuel-related. I determined INCORRECTLY that it wasn't due to the injector (it turned out it was), which led me to rule out the suction valve and the delivery valve and determine INCORRECTLY that it was the plunger.
--Reading further and talking with other folks, stuck plungers may be associated with tiis that have sat for years, but are highly unlikely to be associated with well-running cars. That is, the piston can stick from disuse, but it is unlikely to start sticking in a running car that's humming on down the road.
--I now know that, if a stuck plunger is suspected, the pump head does not have to come off. You can simply pull out the suction valves, which exposes the tops of the pistons, and use a thin non-marring wood dowel or plastic probe to push on the tops of the plungers, which, as per the photo above, are right there. You're not even sticking anything down a cylinder. This was described in another thread, but I didn't understand how it was possible until I saw that the plungers are not pistons. They're not fixed. They ride on springs. That's why you can just push them down.
--You also can pull the belt off the injection pump, rotate the pump pulley by hand, and verify that the tops of the plungers go up and down. They don't move by much -- maybe 1/16" to 1/8" -- but they do visibly move. If you see this, nothing is stuck.
So, that should largely prevent you from doing what I did. But...
If You're Thinking of Pulling the Head Off a KFish Pump, Remember:
--If you have to take the top of the pump off, when you unscrew the four Allen bolts holding it on, the springs on the plungers push the head up. When you lift the head up and out, the plungers will slide out of their cylinders. If you're lucky, they and the springs will drop down into the bottom of the pump. It's not that hard to fish them out. But if you're unlucky, they'll go over the side of the pump and fall down in the engine compartment. Not good.
--I don't know enough to know if the plungers and springs are matched for individual cylinders (or matched for delivery and suction valves), but when they fall out, you lose the correspondence of which plunger went into which cylinder, so it's really best not to let it happen in the first place.
--When you reassemble the pump and put the head back on, you can try putting a dab of Vaseline or some other gas-soluble grease in each of the cylinders to hold the pushrods in. Or you can curl four fingers around the right side of the pump head and put each finger at the bottom of a plunger, essentially holding the plungers in place while you manuever the head into position. I had to do it a few times so I got pretty good at it. It actually works pretty well.
--Knowing this, with practice, when you pull the head off, you could probably undo the bolts, let it rise up from the pressure from the springs, reach around the right side with your fingers, and get them in there and hold the plungers in place.
Like I said, I learned a lot...
My son and I completed the installation of an electric fan in our 1970 Chamonix 2002. As with most things with this car, it wasn’t particularly difficult. It did require a little research (on this site), access to a couple of wiring diagrams, a small pile of parts,and a modest level of problem solving skills.
There is an excellent “how to” article by Zenon Holz at this link http://www.zeebuck.com/bimmers/bmvseite/)and a great wiring diagram from Shaun and Zenon at this one: Electric Fan Wiring He obviously has better than average electrical and fabrication skills, and I chose to go a simpler route and bought parts to do the job.
First though, to describe the problem. It was not cooling the engine. That problem was solved with a new thermostat and water pump. I was also bugged by a small dribble of coolant out of the base of the plastic neck on the 320i (?) radiator and I noticed that the tips of the fan blades were worn off from contact with the radiator. So I replaced it with a 3 row Silicon Garage radiator. Temperature ran about 205 degrees on the VDO gauge and showed about 2:00 o’clock on the stock temp gauge. So far so good.
However, one day I noticed that the brand new radiator had been in contact with the fan. Not just wearing off the tips of the new fan blade like it had on the 320i radiator, but lightly marking the radiator fins (see picture).
After talking to a mechanic friend and searching this site, one of the suggestions to a post was to heat and bend the fan blades to provide more clearance between them and the radiator. So I did. Bought a heat gun, set it to 1200 degrees, and bent each blade about .5” back. Voila, more clearance. I reinstalled the fan, took the engine up to 6000 RPM and while the fan blades flexed forward, there was plenty of clearance.
Problem solved? Nope. During spirited driving on the highway I smelled what seemed like burning plastic. Upon inspection, it was clear that the fan had hit the lower edge of the top radiator tank. My assumption was that the fan had flexed forward during one of my high RPM moments and the blade bending process had weakened the plastic, causing it to flex forward into the radiator. So, my next step was to cut down the fan blades with my trusty tin snips , the theory being that the shorter blades could never flex enough, no matter what the RPM, to chew up Curt’s radiator. The amazing thing is that this didn’t affect the cooling ability at all.
For those of you that know moles…they are pesky critters that love to hunt insects and grubs and in the process wreak havoc with your yard. Some people flood their tunnels, some people use poison smoke, but I trap them. I want positive confirmation that the problem is solved. Same with this fan problem.
I assumed that the fan was the problem. Upon reflection, I think it was a small front end collision sometime during this car’s life that closed the gap between the radiator and the fan just enough to cause contact during the right circumstances. Those circumstances I believe now, had nothing to do with hard acceleration and fan flex, but exactly the opposite. I concluded that the contact occurred due to the inertial force of the engine under hard braking while the rest of the car is slowing. This may be due to the stock rubber motor mounts that I installed to get rid of the steering wheel vibration caused by the urethane mounts that were in the car when I bought it. In fact, during the day of spirited highway driving described above, I did indeed jam on the brakes at high speed to avoid a sudden slow down ahead of me.
In any case, I decided that I could either replace the nose to improve the clearance and correct the previous repair or install an electric fan and eliminate clearance as an issue. Seemed like an easier and cheaper solution as well as getting rid of the (unnoticeable to me) power loss from the stock fan. I still ended up with a new nose in the garage attic, but that’s a different story.
So, on with the fan installation. I read all the FAQ’s, Zenon’s article, and ordered some parts:
12” SPAL 1360 CFM pusher fan.
$112 - Fanman part #2030
Hayden electric fan control, variable temp setting, with a probe temperature switch.
$ 36 - Summitt Racing Hayden #3647
Massive mounting brackets
$ 75 - Massive
Manual on/off switch. 30 Amp
$ 3 - Radio Shack
Miscellaneous electrical connectors
$ 3 - Radio Shack
First, I chose the probe style switch because I had nowhere to put one more temperature sender. The coolant neck had the stock sender plus I drilled and tapped the spare boss for the VDO gauge and while I considered the bung at the bottom of the radiator, all the advice said to take the temp at the top, where the coolant was the hottest. The choice then was some kind of inline hose fitting or the probe. I picked the probe.
I also chose the Massive mounting brackets even though they were pricey compared to the other $5 dollar kits but they came with one thing the kits did not: Lee’s phone number. It came in handy and is the reason I am writing this note.
First step was to drain the radiator, remove it, the fan and the front grills. The brackets were then installed on the fan (see pictures) and the fan+brackets slipped inside the nose through the grill opening. First problem: the fan bracket runs into the right side nose supports preventing the bracket from aligning with the upper right radiator mounting hole. Strange thing with Lee’s kit: it came with both screws and clips like the stock radiator mounting set up as well as bolts and nyloc nuts. I used the bolts and nylocs to mount the fan to the bracket and could have jettisoned the stock radiator screws and clips and used the bolts, but I was wondering why Lee included the new clips and screws. So I called him.
It turns out that which was intuitive to the designer was no so obvious to the untrained user. I was trying to mount the fan to the front side of the sheet metal wall, on the other side of which was the radiator. But that nose support was in the way. So Lee’s design really intended the fan to mount on the same side of the wall as the radiator, held in by the same screws and clips as stock, which he had included in the kit. Got rid of the problem with the nose support. Presented another one however.
The 12” part of the SPAL fan dimension is the diameter of the fan blades. It also turns out that the size of the opening in the wall in front of the radiator is 12.25 “, or so. Sadly, the housing on the SPAL fan that holds the fan is 12.5”, a bit too large for the opening (see picture). Now , if the fan were in front of the wall, that would be no big deal. A little surgery on Lee’s gorgeous aluminum bracket and it might bolt right up with the nylocs and bolts provided. However, it would also have left a gap between the fan housing and the radiator causing the pushed air to exit out the sides instead of being forced through the radiator. So, I solved this problem with a little sheet metal modification to the wall in front of the radiator. I cut off .25” on the top with my Dremel tool and cleaned it up with a file. The fan now fit through the hole and the mod was invisible once everything went back together. See pictures.
The fan is now sandwiched between the radiator and the wall using the stock radiator mounting. The fan housing is directly on the radiator so no air escapes while the fan is in operation.
If would have helped if I had thought to insert the probe though the radiator before I installed the radiator. But I didn’t. I wanted the probe and its wire out of harms way so I took out the radiator and installed it from the front. 2.5” opposite the upper left radiator bracket mounting hole. This allows it to the clear the sheet metal wall and not hit the Massive fan bracket.
Last step was the wiring. The Hayden kit came with a variable temp controller, relay, wiring for 2 fans, an A/C wire, the probe and a fused power line to the fan. A couple of comments. The fuse holder has no cover (not perfect) and the kit included minimal instructions, so I still don’t know the difference in temp between fan on and fan off. Testing will tell. Lastly the probe wire was barely long enough to reach the radiator from the spot where I wanted to mount the relay. That problem solved itself. I’ll explain later.
I mounted the fan relay below the horn relay which is the round thing next to the voltage regulator on the left inside wheel well. I picked this spot because it was close to the battery, had a good ground location and the green/blue horn power wire was right there on the relay. I screwed the fan relay to the wheel well, connected the 12 v power feed to the green/blue relay terminal using one of those two into one spade adapters, connected the fused power line to the fan, ran a new ground wire to the fan harness, connected the pink wire from the relay to the battery’s positive terminal, cut off the wires for the 2nd fan and the a/c and was almost ready to go.
I decided to add a manual switch inside the car. I fabricated a little aluminum bracket and used the 5mm x .80 screw hole on the bottom steering column cover to mount it. I grounded it and ran a wire to connect to the radiator probe. I simply used one of those blue crimp connectors that allows you to connect a new wire (from the manual switch) to the existing probe wire without stripping cables.
Great idea it seemed until I discovered that the fan would never turn on unless the manual switch was on, which meant that the relay circuit was doing nothing. It turns out that the blue crimp connector had severed one of the two wires going to the probe. Well, the probe wire was too short anyway, so I cut it, spliced in two new lengths of wire left over from the relay harness (the two probe wires are not interchangeable. Once severed they need to be reconnected to their original twin on the other end of the splice) and then had to figure out which one of the pairs to connect the switch wire to. Looks simple in Xenon’s wiring diagram, but it wasn’t obvious to me, so I tested the manual switch wire on each splice. One of them worked and I soldered the switch wire to it, separated the splices from each other with electrical tape, bundled the splices together then cable tied them into the wiring loom that ran to the headlights and horns.
All of that seemed to do the trick. The fan was set to keep the car at 200 degrees and so far so good. I have not road tested this to see if I can cause any catastrophe, but I left that task to my son, who of course always drives in complete control (just installed a new rim after he “slid on the leaves” into the curb).
I hope this helps the next person that tries this installation. Overall, it took a couple of hours, one phone call to Quebec, a little trouble shooting, but I now think my fan clearance problem is solved.
Someday, I will put the new nose on the car, but that can now wait until the 5 speed and LSD , which are sitting on the work bench, get their turn.