single cylinder, waste spark, random sync and ignition timin

[grippo]

Typically in M - 0 crank installs, there is a projection on the flywheel which is used for the sync. This is welded or bolted on, so it can't move, so there isn't any slippage as there is with a cam due to twisting or chain stretching. But yes, if the crank sync was set in the middle of the tooth, there could be ambiguity, but it would only cause things to be off by 4 deg or so with an 85 tooth wheel. And it would probably be consistent, for example in cranking with a low VR voltage it might detect the sync after the tooth, but as rpm increased, the VR voltage would increase and it would detect just before the tooth. This could be tuned out by accounting for it in the spark advance table.

The Cam_Tooth logic was really meant for missing teeth + cam sync and is a way to filter the cam signal so anything that comes in from the tooth after the gap to Cam_Tooth later is thrown out. It didn't occur to me until your problem that it could as well be used for the M-0 case.

[bergsls]

Well, yeah, the crank sync and trigger wheel, (starter ring gear) are permanently married, as they are all part of the flywheel, the teeth will never change relation to each other, but the sync sensor, (which was the origional trigger for the CDI box), can be moved a little on the backing plate, (that how you adjusted the timing), hence you could change the relationship of how the 2 waveforms "overlap". Maybe that doesnt matter, but read on..
I tried changing the cam tooth to 85, only to be dismayed with very intermitent, and somewhat random spark, although the majority of the sparks were falling at a peticular spot. The "cam sync" was constantly re-setting, 1-2-3-2-3-2-3333-2-3-2-3333-2-2-2.... no trigger fix counts. So I went back to 0 cam teeth, greeted with my origional problem, went back to 85 cam teeth, same senario as before, random sparks, resyncs, and miss-timed.
Ready to go home, I changed the dual park mode to M-0 with FALLING sync, (versus rising sync), and then had fairly even, consistent spark, tried several times again, and looked good, hooked up sparkplug, and started right up, after fiddling with fuel settings, and timing tables, has been running good, and have confirmed many times the timing has been consistantly right. Once in a while will get a little "chuff" out the exhaust when starting, but maybe could be tuned out.
But after a while, and trying to rev the engine, I now get a huge backfire out the exhaust at approx 4000 rpm, I need to get my wideband in the exhaust, but am sure its with rpm signals, the tach gage will drop to 0 instantly when it backfires, but don't see any trigger fix's or a re-sync when that happens. This thing has a turbo, so I instinctly just get off the throttle. This is all in neutral. It's not the rev-limiter, which is soft fuel cut anyway, much higher at 5200 rpm.
Is there any settings I should be concerned about, trying to reach higher RPM? I didn't have more time to really see what is going on, I will get the wideband on it, (I'm sure it's not that), but the dropping Tach gage has me concerned. Do these need resistors on the crank sigs? With a steel trigger wheel, I know this thing is cranking out some volts. I never experienced this before, when I had the origional problem, as I know I reved it to over 5000rpm several times, but so many things have been changed since then, so who knows what is really the same as before.

[grippo]

Trigger on falling vs rising can make a huge difference, sorry I didn't think of that, but you had what appeared to be such a good signal it didn't occur to me. There is a writeup on this in the documents with pictures of pulses. When they cross the 0 volt line with a shallow angle, there is a lot of variation in where it crosses the line, so timing will move with it, but it still can't move more than a tooth and you were seeing a lot more than that.

The other problem that can appear with missing teeth wheels is that a false noise spike appears in between the gap at high rpm. That would be my guess in your case, except you don't have a gap and the teeth are very closely spaced. But at least there is now progress. I would guess that the exhaust backfire may be tied to changing polarity on the VR sensor - that would be the most pertinent change you made in that area. If the tach gauge went to 0 momentarily then that is a reset - it means that exactly 85 teeth did not pass between cam syncs, so the ECU starts all over except it happens way faster now because you are going 10x faster in rpm. The trigger count may have gone up and down and the datalogging isn't fast enough to pick this up.

Do you have a V3 microsquirt (plastic case)?

[bergsls]

Yes, it is the V3, newest one just released several weeks ago, plastic case.
I just wanted to comment about the rising/falling signal, in that changing that parameter only made a difference when entering the "cam tooth" to 85. When the "cam tooth" was set to 0 as it always was priror to a day ago, changing from "rising" to "falling" didn't change any outcome, (erratic timing).
I am a little unclear about what you meant bywhat is causing the backfire and (most likey) resets. Are you suggesting I switch the VR polarity on the 85 signal? We probabally don't have enough info yet on that, and hopefully will get some free time to see what is happening, and maybe get some more concrete info.

[grippo]

I don't know what caused the backfire, and I'm not sure what caused the loss of sync (indicated by rpm going to 0 momentarily). As I said, the high rpm loss we have seen is always associated with missing teeth. and even switching polarity is mostly a help for missing tooth wheels, or wheels with a small number of teeth. The new V3 chip is more robust than the previous design, and again the only problems we have seen are with missing tooth wheels.

What would be of most help is to get some scope shots of both the wheel and sync signals just before it hits 4000 rpm. Without this it is hard to say what to try.

[24c]

...I now get a huge backfire out the exhaust at approx 4000 rpm...

You really need to datalog this to identify the cause, but a reset on my OEM ECU would discharge the coils, but I am unsure whether the Microsquirt does this. To be honest I have never noticed it in all my fiddlings, so I suspect it doesn't. You do notice it with the OEM system, because you get a soft "phuff" through the intake, as the unburnt mix is ignited.

[bergsls]

Got a little time to dabble with it today, and no the fuel is OK, it's re-setting, although never see a trigger fix, or the sync status change, but do notice the tach gauge drop, and the "tach counter" on the black screen always reset to 0 when the backfire/rpm stall happens, It was origionally at 4100 rpm, then 4600 after I adjusted the running pulse tolerance from 25 to 30 %, but any further change made no difference, (maybe just a coincidence) but whatever it is, it's very consistent that at 4600 rpm, you will get a few soft misses, and some tremendously loud BANG B-BANG BANG!
I scoped the crank wheel and saw no abnormal changes at that point, just that the peak to peak voltage raises up to about 3000 rpm, then slowly decrease the higher the revs go, but suspect it is the desighn of the sensor I am using. Even at 4600, it still has 65 volts peak to peak. It is a wheel speed sensor commonly used on GM trucks. I know I know, it wasn't meant to generate the frequency I am asking, but pattern always looks great, (nice sine wave)
I went to the sync signal, and looks ok until about 3000 rpm, it starts to develop a funny "ring back hump" (I call it), and develops more and more pronounced the higher the revs go, untill 4600 rpm, it almost reaches the ground reference. I will post a shot of the scope just at 4600, where it breaks up. Could this be a problem? causing a double trigger on the sync? I only have the typical automotive shop type digital storage osciliscope, and could never get all the events of one rev on a screen and be remotely ledgible, but really dont see the pattern break up, just develops that ring back hump more the higher the revs go.
The sensor for the sync has a funny resistance of only 100 ohms, seemed really odd to me, as I am used to 600 to 1200 for most automotive VRs. The only reason I used it was because it was there, all set up, I just narrowed up the very wide hump it used on the flywheel. See picture. The trigger hump is very clean and consistant for and aft of the sensing element track

[bergsls]

Here are the pics I wanted to post above, but had to reduce the size to post them, so here they are

[bergsls]

OK, those are now junk pics, lets try this

[grippo]

That last picture is what it looks like when a missing tooth wheel runs into this problem at high rpm. That could definitely cause your problem if it makes the triggers occur between T84 and 85 and then again bet. T1 and T2. Then with Cam_Tooth = 85 it would get thrown out in both cases. It would cause no problem with cam T = 0, but we know that messes up starting. But it might work with cam T= T84 and still not mess you up during starting.

BTW, when cam sync causes the loss of sync, you won't see any change in the trigger count on the datalog because this only applies to the 85 tooth wheel. In the loss of cam sync case there are only 2 indicators of a loss of sync: rpm going exactly to 0 (not just dropping a bit) or the sync status dropping from 3 to any lower number. If it happens at high rpm and the resync is very fast, the datalog might not be fast enough to pick up any of this, but if it happened enough times you would see it at least sometimes by luck.