TravisR

iTrader: (9)

I would set them on the high side and leave them. The coils will get warmer, but they will tolerate quite a bit of abuse. It just comes out that our cars use a really long dwell stock that was causing them to fail. 4.5-5ms is just too long! Where as 2.5 would more then likely never cause a problem even if they are being overcharged just a little bit.

y8s

iTrader: (8)

so ideally you want to never hit the current limit. which with my super rough graphical estimation skills would be around 2350 to put you at 10A at 14.5V (assuming a voltage drop in the wires).

of course if they are very tolerant, run whatever you want and see if they heat up and die.

ZX-Tex

iTrader: (15)

I am going to leave it at 2500 then.

Stein

iTrader: (46)

FWIW, I ran my stock coil at 2.5 ms for a about a month until Travis suggested going to 4.5 and it has been there since. No noticeable change in performance. Should I knock down my dwell to 2.5 or something in between?

As an update, I went to lunch and it ran fine. After lunch on a hot start it ran like crap. Wouldn't idle on the first six starts. Died immediately. I nursed it with the throttle and got it to run and drove back to work. It wanted to die at every stop and a large AFR spike at tip in. Towards the end of the trip it at least started to idle again but with a droop and less AFR spike.

I sent a full log of that run to Travis along with this morning's good run and my current ECU file. This was running on my map in closed loop with Matt's closed loop enhancements.

ZX-Tex

iTrader: (15)

From what I remember you should keep your dwell at 4.5 for the stock coils. But, hey, if you fry them, I'll sell you mine

Sounds like we are both still having the same problem then. I am at a loss here.

Stein

iTrader: (46)

I'll keep it at 4.5.

Man, I thought I had it licked when I went to closed loop and it was sooo perfect this morning.

ZX-Tex

iTrader: (15)

Yeah mine ran great this morning too. I think if we have any correlation at this point, it is heat soak caused by stop-and-go driving, or something like it, in warmer air temperatures.

OK I am reaching for something here. From the manual
"The current for the injectors can be set. This would typically be set at 0.9A, although
higher currents can be used if more injectors are installed. The outputs can be set to
peak-hold type or constant current. Peak-hold is preferred, as it decreases the injector
dead time. Some injectors (the standard Suzuki Swift GTi come to mind) require a
higher current (1.5A) to open properly."

Right now I have peak-and-hold on, and current set at 0.9A, per the manual. Travis, Rob, since these are RX7 550cc 'purple top' high impedance injectors, should we try running constant current instead of peak-and-hold, and/or try a different current setting? Do you know of anyone else running a 4-cyl with RX7 550cc injectors on an Adaptronic, and how they have it set up? I would play around with this, but do not want to damage a good set of injectors doing something stupid.

ZX-Tex

iTrader: (15)

OK well it did it again on the way home, a few times. Once again after stopping to fill up on gas, once after stopping at a stop sign. I have a log, but have not looked at it yet.

I am going over a few things:

- I recalibrated the MAP sensor using a Mighty Vac pump and it seems to be reading a bit more accurately now. It is reading 97-98 kPa which is about right at my altitude of 1350 ft.
- I flashed version N of the firmware, which seems to be the latest non-beta firmware version.
- I possibly have a weak battery. It can only crank the car a couple of times, and gets drug down below 12V in only a few minutes when the ignition is on (car not running). I'll see about getting that load tested tomorrow. I have it on a charger right now. Not sure this is the cause, since I still had problems at 14+ volts in cruise. Just including it in the spirit of providing as much info as possible.
- I am also checking for vacuum leaks and what nots.

TravisR

iTrader: (9)

I'm going to be getting some new Adaptronics in, and if this hasn't cleared up by next week when I should have them then I'm going to swap you units and see if that changes anything.

ZX-Tex

iTrader: (15)

Great Travis, thanks.

ZX-Tex

iTrader: (15)

OK I am not certain yet, but things seem to be improving. Here is what I changed:
- Recalibrated the MAP sensor to read more closely to actual ambient pressures.
Lower value 16 kPa 290 ADC
Upper value 300 KPa 2284 ADC
- Found a vacuum leak and fixed it That would explain the high idling. Now the idle RPM stays between 1000-1300. There is still some RPM droop when the loads come on and off at idle but it is better.
- Retuned the idle fuel cells, mostly those below the idle range, like the 0 and 500 RPM columns and the lowest kPa cells. The car does not run there much if at all but I assume there is some interpolation effect when at say 700-800 rpm.
- Put the battery on a 'Battery Tender' trickle charger overnight. It was showing fully charged this morning.

I still had the car die at idle after driving and parking. It seemed to recovere more quickly than usual however. And, based on experience so far, in several instances while I was driving around when I would have expected it to stall and die, it did not. So I think it might be less frequent.

As soon as the car died I noticed the voltage was at 11.1V; I had been watching the voltage right before that and it was OK, somewhere between 13-14.5V or so. So the alternator is charging, but once again the battery voltage drops off a lot after the engine dies. It makes me think with the big fans, stereo with external amp, A/C, lights, WBO2 sensor heater, etc. the stock size battery is not up to the task.

I have also noticed in the logs that there can be some very brief (less than one second) but large voltage drops (like 1-2V) that recover quickly; they are not explained by variations in RPM, and do not happen just at idle, but at cruise. I suspect the Adaptronic is more sensitive to voltage drops and brain farts when this happens. Plus, maybe the way it controls the alternator causes these large spikes. Either it is creating the spike, or is not reacting fast enough to recover when a large load comes on, like the A/C compressor and fans. THIS IS SPECULATION ON MY PART. I do not know this for a fact so do not think I am slamming Adaptronic, just raising the idea.

The voltage variation makes me wonder if one or some of the following might be worth a try:
- Switch back to stock ECU control of the alternator (MS method)
- Convert over to a stand-alone voltage controller or an internally regulated alternator
- Put in a more manly battery, like an Optima, to help with voltage drops
- Put a stiffening capacitor in at the battery. I have a 0.5F 18V one sitting in my garage.

Thoughts?

UPDATE:
- I ran over to Auto Zone; we have one right next to where I am at. The almost exactly 1 year old Duralast Gold battery was bad according to the load test. You should have seen how fast the voltage dropped under load. That battery is gone, replaced with an Optima 34/78 Red Top; it fits just fine, with a quick parking lot mod of the battery clamp. Auto Zone deducted the cost of the Duralast battery from the cost of the Optima since the Duralast had a 3 year full replacement warranty. Sweet.
- Despite the Optima, the car still died while at idle. A/C on, stereo on, voltage around 14V. No LC-1 code. It took 4-5 minutes of idling for it to happen. The coolant temp was at about where I have been having problems. It took two tries or so to restart it, then ran really rough for 15-20 seconds, AFRs everywhere, popping exhaust, but then started clearing up and then idled normally again.

y8s

iTrader: (8)

thoughts on voltage: the alternator's not doing anything when the car dies. the battery should return to ~12V with the car off, shouldnt it?

you saw my datalog with the stock ECU controlling the alternator...

I would verify there is no difference of ground levels between the adaptronic power grounds and the battery negative terminal.

Stein

iTrader: (46)

Some screen shots of my last two logs.

First is first start in the AM after it's obligatory two dies. Never stays running for the first two starts. Third revs and then drops back.

Next is the first stop sign after a cold start. Motor is full hot as it is 1 1/4 miles from my house.

Next is hot start after sitting for lunch.

Next is last stop after a hot start.

Added one more: After idling at stoplight, went to take off and touched throttle and it died. Can see voltage drop.

ZX-Tex

iTrader: (15)

Nice logs Stein. Looks like 5 or 6 restarts on the third plot? That sounds about like what I was seeing before.

BTW see the update above (post #111) on my battery.

Stein are you running a parallel fan mod setup?

y8s

iTrader: (8)

John, zip and post the cold start (1st) and hot start (3rd) datalogs along with the map you were running when you took them. I think we can tune that out.

same with the last one with the throttle tip.

Prospero

In the terms of hardware, voltage is critical. We manufacture a lot of equipment that will die horrid deaths with low voltage to the processor.

Strangely, a few of my other project cars also stall with low VDC as a product of the ECU not doing its job right. Began to run larger batteries and better power cables to the fuse block on one car... got rid of the problem quickly.

Maybe, just maybe, the load rating on the PCB between the factory ECU and the Adaptronic is not as good as it needs to be? Dunno if / how it was tested.

Just my

Cheers,
Prospero

Stein

iTrader: (46)

5 or 6 restarts. Would have been more but I got tired of it and "assisted" with the throttle after it started the last time.

Your post 111 is why I checked them to see what I was getting.

No parallel fan. Only running the normal cooling fan. The big Koyo didn't require it so I left it like stock.

Stein

iTrader: (46)

Sent to your email. I can't seem to ever be able to post anything other than a picture file.

Car is on jackstands, though. In the process of installing the Magnaflow.

ZX-Tex

iTrader: (15)

OK cool. Just looking for similarities between our setups.

ZX-Tex

iTrader: (15)

Try zipping it. Sometimes that works for me.