DIYEFI.org Forum

Just swapped the MAP hose back to where it was and did a small experiment with the new fixture. I'm stoked to say that having both new tubes breathing gives me a smooth self-maintaining cold start (with just extra priming at the moment) and that after it warms up it hits a mere 1700 RPM. Blocking both tubes once warm drops it back to normal. I'll be using this bonnet-lifting technique to warm up the car in the morning before leaving from now on. At some point I'll upgrade to a PWM driven system. I look forward to the day.

Today, mixed bag:

Totally died, key on/key off, multiple times, did not help. Did this twice.

Poked around thinking out loud with em_knaps and happened upon a loose battery ground terminal with some corrosion. Fixed this. Could explain the total death via unfiltered alternator noise if it was coming loose randomly while driving.

On the way home I had:

1) rough running at light load around 2k (possibly too rich or too lean?) didn't get a log
2) consistent sync loss on high beam on/off, likely from relay spike undamped with no diode
3) rough running with heater-fan on high - didn't get a log.
4) a couple of deaths by clock loss and resets - did get a log
5) radiator fan seems to have died, hit 115C while idling parked before noticing. No coolant leaks, drum tight hoses. Driving sorted it out with airflow.

I don't mind consistent problems, it's the intermittent/unpredictable stuff that's annoying.

So my plan is:

1) Fix radiator fan or replace with ECU controlled fan setup Fixed itself :-/
2) Replace dizzy wiring with shielded cable Was already shielded, now slightly shorter.
3) Possibly shorten/tidy loom and/or wideband wiring
4) Possibly add diodes to factory relays

This should leave me with only BAD problems. Sync loss should be a thing of the past. Passenger (wife!) will be more happy with more leg room, less wires to worry about.

I may try simply running a bulk capacitor on the power/ground feed. I may trace it and see if it makes sense, too.

Fred.

Today I:

1) Removed dizzy
2) Fixed some iffy wiring that was going to cause a breakdown in future
3) Secured the fitting in the housing
4) Made the nuts that hold the hall sensor captive (soldered them to the steel plate)
5) Shortened the dizzy wiring by about 500mm and confirmed proper shielding and connections (all mint)
6) Cable tied the wideband and wiring for it away from the sensitive stuff

Then I went for a drive and caught the weird rough stumble in a log! :-D What was it? Chronic sync loss! :-/ Why? Not sure. But at least I know. BRV is DEAD stable during these shenanigans. All other ADC inputs are also stable. No correlation at all. Could be masked by excess hw filtering? But... I doubt it.

Today I:

1) Set about installing suppression cap on the coil
2) Attempted to start engine, failed, was confused
3) Disconnected cap, tried again, similar, wtf
4) Got it running, but would die under WOT
5) Died all together! Coil failed, again! :-o This was probably the reason for the random noise and sudden deaths, especially in light of definitely having COP code in the firmware.
6) Installed spare coil from behind driver's seat, ran fine
7) Reconnected cap, still ran fine
8) Made a proper bracket for the new coil and removed the old stuff

I'm thinking that a lack of the cap could have caused primary arcing on misfires. That's pretty much the only way modern coils die - internal arcing. Well, unless you run a m$ box, then they can also burn/melt.

As such I expect the new identical coil to last longer and put out less noise. However I'd also like to reduce misfire occurrence in the first place. I get both rich and lean misfires, however lean ones probably have a healthy happy spark. Rich ones, not so much... Rich ones also imply heinous waste of fuel... Thus I'll probably try to fix this first with better warm up code, damped map sensing, and more tuning.

Fred.

Still lots of sync loss. No change. Sigh.

Added a 1nF cap and another 1k pull up and so far so good. Going for a test drive shortly, will log it and report back here with results :-D

No real change, however I think I know what's wrong. Dizzy rotor-to-cap alignment. And before I hit submit I went and fixed that, and it's still losing sync periodically, though it does feel stronger, so I may have fixed yet another issue that wasn't my main problem in the hunt for the main problem.

Fred,
Maybe a dumb question but did you also change the plugs when you changed the coil ?
Plugs can react weird once they had some rich misfires.

I didn't change the plugs, no. The new coil has a faster charge time than the old one and likely draws more current too. Aside from that, it is probably emitting a hotter spark. I read back through this thread and checked my logs, and these issues began right after the coil swap.

Other curious things:

1) I see the tacho jump up a little bit sometimes WITHOUT sync loss.
2) The sync losses primarily occur between 1500 and 2000 when taking off, even if smoothly. Possibly my shitty tune causing it.
3) It's running noticeably smoother and stronger since aligning the cap better.

Next step is to tidy up the coil wiring which has been a bit ugly since the coil swap. I should pull a plug and check to see if they are resistor plugs, or not. However the leads themselves offer some resistance anyway. When I last pulled them out they were clean and looked good, despite the abuse that they've taken.

Any input is totally welcome as I'm running out of ideas :-)

Fred.

So it turns out that my distributor was indexed incorrectly rendering timing substantially too advanced. And in the process I found/fixed a bug and made the world a better place :-D EMStudio decoder offset wizard works pretty well :-) I guess it has for ages, but a slip up made it in. Fixed now. MASSIVE current timing display is awesome for seeing the number from a large distance eg laptop on roof of car, you under bonnet popping head up periodically.

It also turns out that my decoder edge angles were somehow now incorrect and require adjustment. I'll be doing that shortly once I back-port some other tweaks that I've made. I'm not confident that this will help the thing behave itself, however it's necessary regardless. I'm going to slap a big cap on the power supply lines too, as aside from sync loss, I've been getting clock losses and resets which can only really be coming in through this wiring.