Road going ITB setup

[squirt46]

Thinking some more:  engine temperature isn't an issue when doing a cold start; our SDS system has a start cycle so we'll figure that one out  ALso, it's a batch injection system so perhaps we're getting gasoline on the walls for that reason, and maybe it will go away at higher rpm/higher air flow.  Fingers remaining crossed.

[Dudeland]

14 hours ago, squirt46 said:

Well, dang.  Jeff got everything installed (Weber DCOE bodies with the injector hole into the throttle bore opened up; the FASS adjustable fuel pressure regulator installed, etc.).  The car started after a bit of trying, but we've still got puddling (even down at 16 psi)   So, a couple of steps forward; a couple of steps back.  We don't have the radiator installed, so no engine temperature signal.  Try that next;  Also try lowering the passenger side of the vehicle such that the gs puddling in the throttle bores will run into the manifold and the combustion chamber, and we'll get a more accurate/representative signal off the exhaust header sensor.  Wondering if there's some way of getting the injector face closer into the throttle bores as well.  I was hoping that being able to adjust the fuel pressure would resolve the puddling issue, but no such luck.  Now I'm wondering if the injectors don't work consistently at the low pressures.  Rather disappointed, and am running out of ideas.

 

In the FWIW department adjusting the pressure lower isn't going to fix the problem.  You need to shorten the injection pulse.  Injectors are designed to run at a certain pressure for proper atomization. 

 

[Dudeland]

I just did some backreading.  You will need a cam sensor.  The reason is that is the only way you can run sequential injection with it.  To help with the pooling you will want to fire the injector with the valve open.

[jimk]

Full batch injection with all four injectors firing simultaneously is something done 15 years ago.  Semi-sequential with 2 and 2 firing half fuel each revolution is next best to full sequential but still pulses on a no air flow into a runner. 

Batch also causes bit pressure plulses in the fuel system and some add pressure snubbers in the piping like BMW did in the early FI systems.

No wonder that the fuel is pooling.   Either move the injectors after the throttles or hang the SDS on the wall.

Edited April 23, 2023 by jimk

[squirt46]

The Simple Digital System is strictly a batch firing one, and not sequential. 

I originally wanted port injection, but couldn't work out a way of trying for the 'stealth' look (with the DCOEs acting as throttle bodies) and fitting the injectors in downstream.  I've got another manifold that might have enough space between the flanges to fit injectors in, but I'm quite sure that such an installation would interfere with the starter (can't

remember which starter I've got, but it's definitely one of hte smaller ones, trying to give us clearances). 

Regards throwing the SDS ECU out - umm, no.  Economics are a factor and it's bought and paid for, plus the chaps that built the company don't live very far from me. 

Regards injectors needing a certain amount of pressure to operate:  I've been trying to find out what the minimum pressure the injectors we have required, but no luck thus far. 

Future options:  for $ 2200 Canadian I could buy a pair of Jenvey Heritage DCOE port injection DCOE style bodies.  Possibly go with Megasquirt, but I don't know enough about their system to know if it would for throttle body; abandon the whole circus get the original Mikuni side drafts working properly or build up a pair of Weber DCOE carbs.

Currently thinking that we should get the cooling system installed again so we can run it longer; get the tach signal working; get the engine up to operating temperature and then start working on 'tuning' the system (inlcuding the fuel pressure) and see how it looks at higher rpms.  

I might be just over thinking this, as I wouldn't even know we had a 'problem' if the air cleaners were on it. 

[2002iii]

2 hours ago, squirt46 said:

The Simple Digital System is strictly a batch firing one, and not sequential. 

I originally wanted port injection, but couldn't work out a way of trying for the 'stealth' look (with the DCOEs acting as throttle bodies) and fitting the injectors in downstream.  I've got another manifold that might have enough space between the flanges to fit injectors in, but I'm quite sure that such an installation would interfere with the starter (can't

remember which starter I've got, but it's definitely one of hte smaller ones, trying to give us clearances). 

Regards throwing the SDS ECU out - umm, no.  Economics are a factor and it's bought and paid for, plus the chaps that built the company don't live very far from me. 

Regards injectors needing a certain amount of pressure to operate:  I've been trying to find out what the minimum pressure the injectors we have required, but no luck thus far. 

Future options:  for $ 2200 Canadian I could buy a pair of Jenvey Heritage DCOE port injection DCOE style bodies.  Possibly go with Megasquirt, but I don't know enough about their system to know if it would for throttle body; abandon the whole circus get the original Mikuni side drafts working properly or build up a pair of Weber DCOE carbs.

Currently thinking that we should get the cooling system installed again so we can run it longer; get the tach signal working; get the engine up to operating temperature and then start working on 'tuning' the system (inlcuding the fuel pressure) and see how it looks at higher rpms.  

I might be just over thinking this, as I wouldn't even know we had a 'problem' if the air cleaners were on it. 

Most modern fuel injectors are meant to run with at least 43.5 psi some newer cars with return less fuel systems use higher pressure.

 

I would suggest getting some properly sized injectors.

 

Definitely put the cooling system back together and get it warmed up, Tuning a cold engine is a waste of time.

[Son of Marty]

39 minutes ago, 2002iii said:

Tuning a cold engine is a waste of time.

Bingo.

[squirt46]

Most modern systems are port injection, and use the 43.5 psi etc.  Throttle body injection (as used initially by Detroit as they explored this Brave New World in the 80s and early 90s) was lower, per my brother ( a grey-haired retired auto mechanic):   

- fuel pressure:             - carburetors:  5 to 6 psi.; 7  ½ psi absolute max., above 8 pushes needle off seat

            - tbi:  9 to 13 psi            (injectors above the throttle plates)

            - port fuel injection:  35 to 70 psi            (injectors below the throttle plates)

We're struggling with a throttle body (TBI}  injection system, so my 16 to 18 psi pressure might be too high (per  the appropriate anciente Detroit wisdom).  Regards 'tuning a cold engine":  I agree.  We're just trying to get it running, identifying problems, and then optimizing. 

[Tommy]

42 minutes ago, squirt46 said:

Most modern systems are port injection, and use the 43.5 psi etc.  Throttle body injection (as used initially by Detroit as they explored this Brave New World in the 80s and early 90s) was lower, per my brother ( a grey-haired retired auto mechanic):   

- fuel pressure:             - carburetors:  5 to 6 psi.; 7  ½ psi absolute max., above 8 pushes needle off seat

 

            - tbi:  9 to 13 psi            (injectors above the throttle plates)

 

            - port fuel injection:  35 to 70 psi            (injectors below the throttle plates)

 

We're struggling with a throttle body (TBI}  injection system, so my 16 to 18 psi pressure might be too high (per  the appropriate anciente Detroit wisdom).  Regards 'tuning a cold engine":  I agree.  We're just trying to get it running, identifying problems, and then optimizing. 

Are you defining your home built system as "TBI" based on the injector positioning? And that is why you are determined you want lower fuel pressure? Like mentioned several times the needed pressure depends on the injector type. If they are port injection injectors they don't just mysteriously start working with 1/5th pressure by moving them 2" on the other side of throttle plates. You need to know to which range of pressure your injectors are meant for. 

 

Earlier you wrote "Looks like one problem is that the injectors are rated to move 31 lb/hr at 43 psi;", so it sounds that you have injector data that gives flow rate at 43psi (3bar). That clearly tells that the pressure needs to be around there. I wouldn't go lower than 2,8bar...that is 40 psi, or higher than 3,8bar, 55psi. So just set the pressure to 43psi and let it be there. Then use your ECU to adjust amount of fuel.

 

[Dudeland]

22 hours ago, squirt46 said:

Future options:  for $ 2200 Canadian I could buy a pair of Jenvey Heritage DCOE port injection DCOE style bodies.  Possibly go with Megasquirt, but I don't know enough about their system to know if it would for throttle body; abandon the whole circus get the original Mikuni side drafts working properly or build up a pair of Weber DCOE carbs.

 

 

I looked at the Jenvy option.  The challenge is that you need to figure out a cold idle circuit, which requires a bit of additional hardware to get it to work with an M.S.  There is a person that used an Elite 1500 (or 2500), but without the cold start circuit.  He lived further south, so cold days really aren't that cold, and he just put his foot down a bit until it got warmed up. 

 

That being said, last time I priced it out, you are looking at about $6-$7K CDN to put a Jenvy kit on in Canada, not including linkage or a manifold.   

 

I couldn't get a complete answer on which type of manifold to use and Jenvy wouldn't recommend one.  Some manifolds share a port between the intakes ports and others don't.   This somehow helps with intake pulses, I am not smart enough to understand what difference that makes.  Others on this forum shared their knowledge and I was leaning toward manifolds that share a port between the intakes. 

 

The MAP sensor built into the Jenvy intakes would be impacted by the choice, although I am not sure how. 

 

There is something that needs to be done to make sure you have enough vacuum to operate your brake booster as well. 

 

P.S.... yeah pressure needs to be at the manufacturer's recommendation,  to get less flow, you need to re-size to one that flows less if shortening the injector pulse width doesn't work.  You can cheat a bit by turning down the injector pressure, but not much. 

 

With my Sniper EFI throttle body, I could tweak the minimum pulse width.  From what I can tell you can adjust the pulse width with your SDS, but it isn't intuitive. 

 

Here is the table to help with injector selection taken from the SDS manual (EM5) 

 

http://www.sdsefi.com/em5manv28.pdf 

 

 

 

 

 

 

 

[d.hitchcock]

Data point, from someone running Jenvey ITBs with no cam sensor (batch fired), and no special provisions for cold start/cold idle.

Cold weather start (below 40F) requires a couple-three turns of the key to get the car to fire. I have to keep idle up (heel-toeing the throttle if needed) until some temp shows on the coolant gauge. Then the car begins to run and idle more or less normally.

The air-gap for the crank position sensor seems to be very sensitive to big changes in ambient temperatures. And the car's ability to start depends on that gap being adjusted properly. I check it every oil change now.

 

[squirt46]

Mr. Hitchcock:  what part of the world do you live in (thinking about temperatures).  If/When we get our project running, it will be a summer only vehicle, so unlikely to be below 40F (we live in Calgary, Alberta, western Canada).

 

[squirt46]

Definitions:  my understanding (and use of terms) is that throttle body injection means the injector is in the throttle body, and injecting upstream of the butterfly.  To me, port injection has the injector downstream of the butterfly.  I believe that this is consistent with the Standard Wisdom, as in Wikipedia:  "Single-point injection (also called 'throttle-body injection')^[17] uses one injector in a throttle body mounted similarly to a carburettor on an intake manifold. As in a carburetted induction system, the fuel is mixed with the air before entering the intake manifold." 

  " 

[cda951]

@squirt46 I am sorry to state the following, but I think you would be better off either going back to carburetors or biting the bullet and using the Jenvey option if a "stealth" Weber-look setup is the primary goal.

 

As others have stated, the injector placement/fuel puddling will always be an issue----perhaps with many hours of fiddling and tuning you could get it to run well, but with such a compromised setup you are negating much of the advantage of going EFI in the first place, which is much easier cold-starting, better fuel economy and flexibility, among other things. OK, one advantage of the fake Webers is you can eliminate the Venturis and get more potential for top-end power, but that is only possible if your engine can flow enough air to take advantage of this.

 

Proper atomization of the fuel is key---carburetors don't do such a good job of this with the idle circuit (which is in play up to around 3000 RPM with Weber DCOEs and the like), but as the main circuit becomes operational, the ingested fuel is focused by the venturi effect and atomization is quite good . . . . . actually can be better than some lower-PSI port-injected EFI systems.

 

In other words, one can waste a lot of time and money (time=money in my world, and I speak from experience) if building a system around a certain set of parts, but that may not achieve the desired outcome . . . .

Edited April 25, 2023 by cda951

[squirt46]

Calculations:  the Wikipedia entry for the BMW 2002 identifies stock hp as being 99 hp for the standard model; 118 hp for the dual carbed Ti, and 140 hp for the mechanically fuel injected 2002.  The turbo had 168 hp, but we're just using a normally aspirated engine. 

The standard formula for sizing an injector (in lb/hr) is horsepower * BSFC/number of cylinders/duty cycle.  Duty cycle (in the material I've read) suggest 70% to 80%; BSFC is defined as the pounds of fuel used per hour divided by the horsepower.

The actual fuel delivery will be a function of the pressure, and I used: 

advertised flow rate * the square root of (the pressure I'll use divided by the rated pressure)

Obviously a non-linear relationship!

Among other sources, I used summitracing.com and holley.com.  And I compared my results with the SDS table provided by Dudeland.  Plus I'm going to compare my stuff with HP Books 'Electronic Fuel Injection' by Ben Strader.