The Project Pad

Ended up rotating the dizzy to match the hash marks with the location of the trigger pin (not sure how it was running before!) but I'm still not seeing signal :banghead: This is looking more and more like a size of the boat issue since I am, now, confident that everything else is correct.

Does it run on motronic?

It used to; no way to test that now #notplug&play

Anyone have a good way of telling wives that the car is not fixed yet because out of a 3 day weekend you've only snuck a few hours of work in? How about explaining that you can't estimate how long it will take because you have not fixed it yet?

I am all ears on that one. My car has not moved all weekend and is parked in the middle of the garage. Out of the three projects I had planned on getting through, I got exactly .4 of one project done..

I was chasing a no-spark issue off and on for a couple of months and finally figured out I had (and possibibly still have) a bad connection at the ECU. Each time the issue has come back, it goes away after unplugging the ECU and plugging it back in. I had it in the service position twice tracking it down (I cracked the corner of my euro coupe bumper one of those time). Luckily, she still had room to park her car in the garage while it was sitting there.

I would just chalk it up to the age of the car. You start tearing into a 22 year old car and somethings bound to break when you put it back together. My wife seems to understand that one.

I would say your problem is probably the timing pin size. The pin is roughly 60deg btdc, so it won't line up with the TDC cam mark. The width of the AAN trigger is about 1/4" square peg, the pin on the 7A and 3B are more like 1/8" and round. You will get a much smaller and lower voltage signal from that then the signal produced by the AAN flywheel. Dave made a cool square plug that he drilled out and slid over the original pin, but spacing is critical, too far out, you smack the sensor, to far in, not enough signal. You can install the pin sleeve thru the starter hole, so at least the trans doesn't have to come out.

On my way to grab an oscilloscope and stock for pin girth enhancement.

No measurable signal on the scope, time for metallic pin enhancement; wish I had a drill press!

Take pics, I'm interested in how this plays out. Good luck!

I'm basically emulating what Hank and Dave did for Dave's car, just using some square stock with an offset hole slid over the pin and welded (I'm tempted to JB it lol).

So pin enhancement, I drilled out a chunk of 1/4" square stock but the best I could manage in my ghetto garage was ~0.4mm crooked from side to side (centered at one face but 0.4mm off center of the opposing face).




Will this work or should I be concerned about a crooked pin? I can grab a drill press, just trying to save an hour of driving (I know, I know).

Mcstiff wrote:So pin enhancement, I drilled out a chunk of 1/4" square stock but the best I could manage in my ghetto garage was ~0.4mm crooked from side to side (centered at one face but 0.4mm off center of the opposing face).




Will this work or should I be concerned about a crooked pin? I can grab a drill press, just trying to save an hour of driving (I know, I know).

Forget a drill press, you need a lathe if you want the hole dead center.

From my own experience in decoding the TDC ref pin on those 135 tooth flywheels, the positioning of the pin is pretty important as it needs to occur squarely between the two flywheel teeth for a reliable decode.

with vems you can set it to trigger at the begining or end of the pin, and tell it how many degrees off TDC it is, that to me means it doesn't have to be super accurate, you just have to dial it in after the fact.

Hybrid_Hatch wrote:with vems you can set it to trigger at the begining or end of the pin, and tell it how many degrees off TDC it is, that to me means it doesn't have to be super accurate, you just have to dial it in after the fact.

Oh but it does need to be accurate since if it occurs too close to either of the two teeth it is in between, noise on the signal will cause one to proceed the other signal in a spurious way with the result being that your timing error moves 2.667 degrees (one FW tooth time) when it happens.

Its problems like this that make the TDC reference pin a short lived method of triggering ECUs.

that pin is ugly

my2000apb wrote:that pin is ugly

Ouch. To be fair I was actually not shooting for center, one side needs to be offset to clear the VR.

TheArchitect wrote:

Hybrid_Hatch wrote:with vems you can set it to trigger at the begining or end of the pin, and tell it how many degrees off TDC it is, that to me means it doesn't have to be super accurate, you just have to dial it in after the fact.

Oh but it does need to be accurate since if it occurs too close to either of the two teeth it is in between, noise on the signal will cause one to proceed the other signal in a spurious way with the result being that your timing error moves 2.667 degrees (one FW tooth time) when it happens.

Its problems like this that make the TDC reference pin a short lived method of triggering ECUs.

I'm not sure I'm not sure I'm following. If you are talking about the teeth that the separate speed VR is sensing I don't see how it would be a problem since this sensor is not currently seeing anything (teeth or pin) and <1v signals are, as I understand it, ignored. There is a good chance I'm missing something

Mcstiff wrote:

TheArchitect wrote:

Hybrid_Hatch wrote:with vems you can set it to trigger at the begining or end of the pin, and tell it how many degrees off TDC it is, that to me means it doesn't have to be super accurate, you just have to dial it in after the fact.

Oh but it does need to be accurate since if it occurs too close to either of the two teeth it is in between, noise on the signal will cause one to proceed the other signal in a spurious way with the result being that your timing error moves 2.667 degrees (one FW tooth time) when it happens.

Its problems like this that make the TDC reference pin a short lived method of triggering ECUs.

I'm not sure about that. If you are talking about the teeth that the separate speed VR is sensing I don't see how it would be a problem since this sensor is not currently seeing anything (teeth or pin) and <1v signals are, as I understand it, ignored. There is a good chance I'm missing something

The ref pin is what tells the ECU the point where to begin counting flywheel teeth for its 360 degree reference.
If you look carefully at a 5 cyl flywheel, you will notice that the positioning of the ring gear puts the tooth between flywheel teeth because you must have a deterministic order in the ref pin VR sensor and the flywheel teeth VR.
Any error between the two from ideal reduces the noise immunity and effects the true accuracy of the ignition timing (be it VEMS, factory, 034 or otherwise). If the ref pin is supposed to lead the next flywheel tooth and suddenly it doesnt due to noise on either VR sensor, you will be off by 2.667 degrees for that 360 degree interval.

Trimming the pin to actual TDC is the easy part and a forgone assumption also in each ECU case.

Also, at low RPMs, you probably will only get a few hundred mv/pp (not volts) from the sensor.
Put a scope on the sensor to know for sure if its working.

VR sensors use a technique called "zero crossing" detection and the actual trigger will be on the point of greatest slope of the VR waveform at the zero volt point. IIRC its the rising edge on the TDC ref pin because audi did the polarity backwards compared to the flywheel VR sensor.
In any case the zero crossing technique will also use a minimum amplitude to enable triggering.
If the amplitude is insufficient, you wont get an output trigger from the VR amplifier.
I'd be surprised if it required greater than a volt since that could make it a no start condition if the signal amplitude is too low while cranking.

Mcstiff wrote:

my2000apb wrote:that pin is ugly

Ouch. To be fair I was actually not shooting for center, one side needs to be offset to clear the VR.

ok touche, i didnt know you were going for that, i still be concerned about it being off center lengthwise, precision machining FTW

File down one of the sides, I would do the one on the left in that picture. This will center the hole, and also allow you to clear the VR.

This is the good side. I may just redo it, the effort of squaring it up is worth more than the sunk cost.

:edit: So I had to cut it in ~half putting be at a worst of 0.2mm off center. I need to file it into a rectangle to fit on the inner side so I think it's a great success!

Attachment ( 31018 ) : ImageUploadedByTapatalk1338402961.486735.jpg

So it will be 180° from here once I clearance the fat side.

Attachment ( 31019 ) : ImageUploadedByTapatalk1338405175.512277.jpg

Awesome. Keep at 'er.

I want holset vids!

Have to build an extension cord and I'll tack then see if it works!

Build under a bad sign?

So in case anyone is searching for info, the starter housing is the tightest thing to the flywheel. After tacking and checking clearance with the block and sensor I was feeling pretty good so I reinstalled the starter and gave it a go. Unfortunately, there was a noise. The only thing that changed was the starter so out it came and I found not only a custom machined starter (no big deal) and a tweaked pin :( With fingers crossed I attempted to nudge the pin straight; the pin decided that was enough and left the building :banghead:

Soooo, I guess I'll be cleaning up the old weld and seeing if I can extract what is left of the old pin and start over. :dur:

Anyone have a bike I can borrow