TrueChaos

Have posted this over on the megasquirt forums as well. I've been chasing a sync issue that occurs only while cranking, but is to the point where the car is nearly impossible to start. The car is a '99 miata, with an MS3Pro PNP, EFR6258 turbo, ID1050x injectors, and stock coils. Ignition is set to rising edge as per the manual. When I try to start the car, I get lots of sync errors, and it is very hard to start. The highspeed datalogger looks like this:

(See CSV - Attempted Start).

When I crank the engine with the throttle pressed down (so flood clear), I get no sync issues at all - the highspeed datalogger looks like this:

(See CSV - Flood Clear)

When it does start, I don't get any sync errors at idle:

(see CSV - Running)

Strangely enough, I can start and run the car with the ignition set to falling edge (with a different trigger offset), with no loss of sync. From what I've read, rising edge should be used as it's a much more reliable signal - and this matches my experience. When I've been able to get it to start with rising edge and I check the timing with fixed advance and a timing light, the marks are stable. In falling edge, the marks seemingly vibrate or twitch a little bit - which makes sense if the signal isn't as clear, and probably results in some timing variability, exactly what I don't want. I don't know where to go from here - is this a sign of bad coil packs? Any help would be appreciated, and I've attached my tune (really more of a basemap as I haven't gotten the car to start consistently enough to actually tune) as well as csv's from the highspeed logger. Please let me know if there's any other data I can provide.

Icedawg

I had a similar problem with my 2001 and PnP unit. The tuner backed off the timing advance on cranking to a total of around 10 and that fixed it. But my problem only showed up starting well below freezing temps.

TrueChaos

Interestingly, mine seems more likely to start with a more advanced spark than a less advanced one - 15 deg seemed to be 'better' before I started playing with things some more. I'm suspecting a few different things at this point:

• I've noticed that I can get it to start easier with say 5-10% throttle. My understanding from the manual is that this is an indication that there's too much fuel. I *think* that more spark advance helping could also be a sign of too much fuel as well, but I don't know. When cranking, I believe that grabs the leftmost VE values associated with the MAP, as I don't have any rows below 700 RPM. My thought here is that it starts to catch, miss-fires / catches very roughly, and the shaking from that is causing the loss of sync issue.
  • Given this theory, I've set a very very small priming pulse.
  • I also reduced cranking pulse width - I'm able to get the most consistent starts with around 50% PW when at operating temperature. I haven't tuned the VE table at 700 RPM / 85-95 kPA and the VE in this range is 93-96, but my VE at idle (900 RPM / 30-35kpa) is around 80. It's entirely possible I have too much fuel in the VE table here and need to pull some of it out.
  • With these changes, it starts 'better', but still not consistent. If anything, cold starts are better than hot starts (which I think again points to too much fuel, given the idle valve opens further - but WUE is active, though small).
• Idle valve behavior - it's behaving as expected normally, but looking at the logs in detail, it seems I'm going from a cranking PW of ~2.5, it catches, the PW jumps to ~4.5 based on ASE & no longer using 50% of the VE value in the calculation, loses sync, and when it loses sync the idle valve goes from whatever its position was right down to 0. Now I've put more fuel in, and choked off the air, so when it regains sync (which it does quite quickly) things are far too rich and it just dies. For what it's worth, it idles at a PW of around 1.5ms.
• Bad cam/crank sensor, or failing coils. All of which are original to the car, and I've ordered replacements for all 3 - might as well eliminate the parts as the problem.

Next I'm going to start by seeing what happens if I turn off ASE, and if it starts better, probably try reducing the VE in the 80-100kpa @ 700 RPM cells, and see what that does.

Icedawg

The DIY Megasquirt guys suggested I turn off the closed loop idel function while trying to sort the cold start. I thought it was not supposed to be active in cold start. But turning it off helped sort the fueling, timing and ASE functions during initial start, and kept the system from appearing to be fighting itself. Later, after getting it starting better, I was able to turn closed loop idle back on with no problems created.

TrueChaos

All my messing about didn't produce anything notable, so I switched to falling edge. Things work perfectly, with a teeny bit of jitter on the timing, so I'm not going to worry about it. There will be a new damper and trigger wheel in my future eventually I think. Had a remote tuner go through the tune and get everything set and adjusted based on datalogs, and, well, the car runs very well but the VE map is super weird. The following VE map tracks the AFR table very well, and the car on the whole feels good. Idle is smooth, returns to idle well, etc. Currently playing with things a bit to correct a slight lean spike (<0.3s duration, approx 2 AFR over target) on return to throttle from overrun. VE/AFR/Spark tables as follows:







Super flat VE table - lowest of 77, highest of 102. I found the following thread with a similar setup to mine that had similar numbers, but there wasn't a resolution / confirmation of an issue:
https://www.miataturbo.net/megasquir...e-boost-96883/

I'm not going to worry about it as the car runs well, but it's definitely weird. Numbers over 150kpa haven't been tuned, as I'm only running about 7psi right now. Won't be going higher than that till I can book time for a local tune on a dyno.