efi A Cam Position Sensor design - please critique

[Healey3000]

Hi,

 

I've concocted a design for a cam sensor that plugs into the distributor location.  Before I cut any metal, I'd like to get your collective thoughts to make sure I haven't done anything untoward, design-wise.

 

There are three machined parts - shaft, housing and cap.  The rest are off the shelf items (McMaster or elsewhere); a donor gear is required from a 2002 distributor.  The shaft is a light press into the bearing and can be either tapped into place or heat shrunk but there's also a snap ring at the top for added measure.  The cap sandwiches the bearing in place into the housing.  A flanged cast bronze bushing provides lower radial and thrust support.  An oil seal keeps things dry but could be optional, depending on how oil climbs up the shaft.  I didn't want to machine an oil groove along the shaft, due to cost.  External O-ring is at the stock location

 

The max diameter of the housing is 38 mm, as is the height of the cap, so pretty compact overall.  I realize there are simpler ways to achieve cam sensing by modifying a distributor housing but...

 

Thanks!

[TobyB]

Quote

plugs into the distributor location.

That seems like you're really limiting yourself for precision piston position.

If it's in addition to crank position, never mind.  Personally, I would try

to get timing info off the cam gear, but I get that this bolts in...

 

Also, with the small diameter shaft on top, it seems like you'll get a lot of sensor 'scatter'.

As in, the ferrous material will trigger the sensor at less- consistent positions than a larger wheel would.

 

off the top of my head,

 

t

 

[Healey3000]

4 minutes ago, TobyB said:

That seems like you're really limiting yourself for precision piston position.

If it's in addition to crank position, never mind.  Personally, I would try

to get timing info off the cam gear, but I get that this bolts in...

 

Also, with the small diameter shaft on top, it seems like you'll get a lot of sensor 'scatter'.

As in, the ferrous material will trigger the sensor at less- consistent positions than a larger wheel would.

 

off the top of my head,

 

t

 

Hi Toby,

 

Yes, this would supplement a crank sensor.  I have one of Tom's trigger wheels at hand.  I wondered about triggering consistency but since it's just for cam position, and most ECU's only look for an edge within a pretty generous window I think it will be okay.  I do plan on doing a test when those Cherry Hall sensors show up in the mail.  If necessary, I have an alternate version with a larger rotor attached.  Same basic layout, otherwise.

 

Thanks for the thoughts.

[burndog]

I love the effort, thought and design, but I am using a stock 2002 Bosch 009 dizzy with Pertronix points for CAM position. Just sayin'... make your life easier. Alternately, you could augment your design with AEM's universal CAM position sensor and mount it directly to you dizzy shaft design for additional simplicity and utilization of your design.

[jimk]

It'll work ok.  You may have to do some trial to find the best gap for reliable signal.  The gap can be sensitive to the tooth/materials etc.  It may have to be jigged up on something like drill press to do the signal study.

Signal scatter is unfounded concern because the cam sensor is not intended to sense position of the cam but rather the signal is used to simply send a cycle reset signal.  The cam signal can float all over the place and the float is meaningless.

When the crank comes around to the missing teeth and a cam signal has been received, the ECU resets the cycle and firing/injection count starts at the beginning.  On the second revolution of the crank the cycle is NOT reset because there hasn't been a signal from the cam sensor.

Be sure the top bearing can be changed easily because bearings can suffer from infant mortality even though most are without trouble from the get go if not damaged during installation.

Edited June 10, 2016 by jimk
Another thought happened

[Preyupy]

The resolution with a diameter that small will be a problem along with the slight "gear lash" at the back of the camshaft (at that diameter the 0.002 thou lash will wind up looking like 10-15 deg of variance, remember it is at 1/2 crank speed so double the error)  .  Why not just put a magnet in the cam gear and mount the sensor in the front cover?  I used a AEM EPM (engine position module http://www.aemelectronics.com/?q=products/ignition-components/engine-position-module-epm ) and it gives me both crank and cam position (2 internal sensors one is 24 impulses/revolution and the other is 1 impulse/revolution.  I modified a distributor housing to drive it not that much different than what you are looking at to do your mod. 

[Healey3000]

Hi,

 

I agree that the small diameter will impact resolution.  However, based on the fact that this is only needed for phasing of the four-stroke cycle, even a backlash as large as what you mention will be acceptable, I think.  I'm planning on using MegaSquirt, and it only looks for a transition once per engine cycle.  Anywhere after TDC, as Jim says, will be fine.

 

How's timing chain lash affecting your AEM sensor for crank position?  I was going to do something like that for my Healey but the jitter would be too great.  Putting a crank trigger on that car is proving to be tricky.

[Healey3000]

2 hours ago, jimk said:

It'll work ok.  You may have to do some trial to find the best gap for reliable signal.  The gap can be sensitive to the tooth/materials etc.  It may have to be jigged up on something like drill press to do the signal study.

Signal scatter is unfounded concern because the cam sensor is not intended to sense position of the cam but rather the signal is used to simply send a cycle reset signal.  The cam signal can float all over the place and the float is meaningless.

When the crank comes around to the missing teeth and a cam signal has been received, the ECU resets the cycle and firing/injection count starts at the beginning.  On the second revolution of the crank the cycle is NOT reset because there hasn't been a signal from the cam sensor.

Be sure the top bearing can be changed easily because bearings can suffer from infant mortality even though most are without trouble from the get go if not damaged during installation.

Hi Jim,

 

Thanks for the reassurance.  Yes, the bearing will be very easy to remove (it's a slip fit).  I was inspired by the design of Nissan's CAS optical distributors and while they use two screws to retain the bearing, I figured on saving some diametric space by letting the cap do the job when the three screws are tightened down.  I may need shims to account for machining tolerances but shims are readily available from McMaster.

[Preyupy]

I have found a harmonic in the chain/cam/gear assembly that manifests as a <2.5 degree variance at 3800-4100 rpm below that rpm it is <1.25 deg and above 4800 it falls to <1.0 deg all the way to 8300. I am running a very light rotating assembly (9.5 lbs off the crankshaft and 4.5 lb lighter flywheel/clutch assembly) along with a 336 cam and heavy valve springs so the "over center" surges on the camshaft are really big, I'm amazed it is as stable as it is.  I very seriously considered a crank mounted sensor but wanted to try this first, I did not need to modify anything on the engine and this entire system can be moved to another engine by just unbolt inc it and moving I to the new engine ( even a bone stock one) 

[Healey3000]

20 hours ago, burndog said:

I love the effort, thought and design, but I am using a stock 2002 Bosch 009 dizzy with Pertronix points for CAM position. Just sayin'... make your life easier. Alternately, you could augment your design with AEM's universal CAM position sensor and mount it directly to you dizzy shaft design for additional simplicity and utilization of your design.

Thank you kindly!

 

I was not aware of the AEM device but it would be overkill for this application since all I need is cam positioning (plus, it's expensive).

 

Interestingly, I made something very similar for the Healey.  Using a Nissan SR20 CAS, I turned down the outer housing to fit.  I also made a custom laser-cut wheel to replace the stock one, with 72 teeth and a single cam trigger.  Jitter on the distributor drive was a lot worse than I had expected, unfortunately.

[jimk]

If you are cheap, for a hall effect sensor, an alternate can be an E36 88-96 cam sensor, pn 12141740383.  I retrieved one from a salvage yard car.  Take the engine harness connector and some wire so you will have both sides of the connector.   It is located on the drivers side of the head, up front.  Tools needed are a big channel lock to take the vanos actuator motor out of the way,   Metric allen wrenches for the socket head cap screw holding the sensor on.  This sensor physical size is near the Cherry and Honeywell GT101

Edited June 11, 2016 by jimk

[Healey3000]

Hi Jim,

 

I did consider that option.  The e46 exhaust sensor is actually very easy to remove, as there is no VANOS plumbing to get past.  It's right up front - remove the air duct and that's it.

 

As it turns out, the cost of one from the junkyard was about the same as the Cherry unit, given the sale Newark was running.  Granted, I now have to find the connector.