Long story short, I just got my '90 TT IROC-Z back on the road after being 'unhistoric' therefore 'unlegal' for the last three years.

All was going fairly fine for a few weeks, then my heater core sprung a leak. On my way home from work, the car went into limp-home mode (which barely works because it's not a stock setup). It died and wouldn't restart, so I went to check the connections to the ('730) ECU, which is snugly wedged into the place near passenger side kick panel (stupid chip tuning, I need an emulator). That's when I noticed the heater core leak, coolant puddled in the footwell. Hrm, but the ECU was in a dry location and had no signs of being wet anywhere. So, I re-seated the connectors, no start. I smelled fuel so gave it WOT, then it started, but chugging badly.

I limped home by pedaling the throttle (and feathering the darn grabby Spec stg III), fortunately it had died only around the block.
The next day I rode the ZX-10R to work because my truck also broke down recently and it snowed on the way home (ice formed on my jacket, too). Niiiiice...
Went to start the Camaro after fully ridding my 1000cc sportbike of the salt and grime (in the snow), and the car idled fine. Hrm. Started it later, started fine. Well... now to get the heater core replaced (I'd dried out the footwell).

Fast forward to today, two storms, 3 feet of snow, and 4 days off work later, and I went to go replace the heater core, but started the car first just to check, as I had between the storms a few days ago. Back to limp-home mode as it was between the storms again, getting SES light as before. Checked the codes that other day, none except the normal code 12, today it wouldn't give me any codes. Hrm. Just to make sure I was right about limp-home mode, I pulled the chip and started it. Runs exactly the same. I re-seated the memcal adapter and re-installed the chip, still limp-home.

So, we've determined limp-home mode, and what seems like dodgy ECU connections (or intermittent shorts) somewhere.

I then read the .bin in TunerPro for the heck of it, and all was wacky, none of the numbers made any sense. Hrm. I loaded up the .bin that it should have on it, verified, and all was well. I then installed it into the car, and limp-home mode still. Came back and read chip, still good. Did that two-step once more to be sure, same results.

So... Is my ECU fried, connections bad, or chip gone somehow? Unfortunately I sent the other three chips I have as well as the other '730 ECU off to New Jersey to be put into my turbo Fiero that's being built (and I'm designing). DOH!
I'm thinking the obvious, the heater core leak shorted something out. What has me puzzled is that it ran fine a few times, it doesn't make a lot of sense.

I need this dumb car to get to work, it's cold (and dangerous) riding in the snow when it's 25 degrees out! (hell, at least you can say I'm a 'real' rider, heh)

Any ideas? Thanks for reading.

 

If you can flash codes then the ECM is not in limp mode. The other test for a running ECM is to key-on, engine-off and observe the SES light. It should turn on briefly, blink off, then turn on solid. Look for the blink off. If it does that then the ECM is not in limp mode.

Can repeat the test after a key-off for at least 10 seconds.

Running rich may be the CTS is bad. Or the connector to the CTS is bad. This makes the ECM think that the engine is colder then it really is and dumps additional fuel in.

It is also possible that there is a mechanical problem. Such as a FPR that has a hole in the diaphragm (dumps fuel into the intake via the vacuum line). Or water in the distributor cap.

As for the ECM and the leaking heater core. Can't hurt to pop the cover off the ECM and make sure that it is dry on the inside. Although the conformal coating usually negates issues with it being wet.

RBob.

 

Thanks a lot, those are exactly the kinds of ideas and info I was looking for.
I'll check them as soon as I get a chance, hopefully we don't have work again because of the snowstorms, but I'm thinking we will, in which case I'd better fix it on this upcoming 3 day weekend.

An initial inspection of the engine bay showed nothing out of whack, but it's extremely hard to see anything through all that turbo junk, and problems aren't usually that obvious anyway.

I think it was only flooded, and not rich, from me trying to get it to start by tapping the throttle (as if it's carb'd, heh) while cranking. Once I smelled the fuel I gave it WOT and it fired up (sort of). I suppose this could mean more than I'm thinking, but I think it was a simple case of "I'm dumb."

FWIW, it flashed codes (well, just 12) the first time I tried, but it wouldn't do anything the second time. I thought that was odd.

 

Also when in limp home mode don't
the fans run all the time

 

That might have worked except that the previous owner wired the fans to run all the time anyway, even in the 'accessory' position. I haven't gotten around to fixing that because I do want the fans to run all the time, but only in the 'run' position. There's a lot of turbo heat under there.

Still haven't had a chance to get to it, I did work yesterday. After my appointment today I'm going to troubleshoot further.

 

RBob,
The SES light does not come on at all with key-on, engine-off. I don't recall whether or not it worked properly before the problem.

It does, however, come on solid when I start the car and keep it "running" by blipping the throttle. It sounds and feels like it's running on about 3 cylinders. Same as before, still no retrievable codes, and same as pulling the chip altogether.

I haven't gotten to pull the ECU cover yet, about to leave for my appt. I forgot that the board was conformal coated, the satellite electrical ground-support equipment (EGSE) I usually build isn't coated so I don't think about it much.

 

Sounds like the ECM died.

RBob.

 

That's what I was thinking, so I pulled it out and removed the G1 adapter, etc. I then noticed some wetness on the board, and also noticed a poor conformal coating job, probably because this was a reman'd ECM.
Rather than even messing with drying it because it could have weakened some of the ESD sensitive components if they weren't already ruined, I just bought another reman'd unit from the parts store (since it's way too cold to go out into the local JY, and since I need this car running asap).

I re-installed everything into the new ECM, plugged it in, and tried to start it. COUGH, then nothing but cranking. I looked over and noticed that I had put my chip back in the ZIF socket backwards! Umm, oops.

I came inside, read the chip, and it didn't match the buffer. So I erased for good measure, verified erase, loaded file to buffer again for good measure, programmed chip, verified to buffer. All seemed fine.

Installed it into the car (the correct way...) and it started and ran normal, no SES... for about 15 seconds.

Bam, right back to the 3-cylinder barely running feeling with the SES light. I hate this car (or at least my inherent need to fix it 'til its broke)...
Then I remembered to check the SES light, key on, engine off. Solid SES this time, no blink.

Came back in and checked the chip, didn't match buffer again! Erased, verified, reprogrammed, verified again, installed in car. Still solid SES light, still barely running.

This time the chip still has the .bin loaded properly (so it says).

So, did I fry the ECM again, the chip, or is this all a bad dream? Someone pinch me?

Anyone want to buy a barely-running twin-turbo 350 6-speed '90 IROC-Z in mostly nice condition for $6000? I have more than twice that into the car, of course.

My V6 Fiero is going to be faster anyway.

 

After another half a day of wasted time, my best theory is that I have a bad chip. Can they sometimes not 'hold' a tune at random? There were definitely some ESD events happening randomly last night with the combination of the cold dry air and my jacket rubbing against itself and me.

I pulled the ECM top cover for a visual inspection of all pins, connections and such, but of course if something is fried it's not possible to see unless it's REALLY fried. Everything was fine to my eyes, except that that factory board has a poor layout IMO. Nothing that would cause problems in and of itself, though.

I went through all the same mess as yesterday, pretty much same results. The only constant is that it's screwed up one way or another.

I hooked up the datalogging and made an ECM connection when the car was running okay. The values were all normal and expected, then out of the blue the SES came back on and the car started its hissy-fit again and the data froze. I checked the chip, it lost the program. Reprogrammed, still running horrible, but 'held' the program this time. Couldn't recover an ALDL connection. Basically all the same BS as yesterday, just with some pointless ALDL data now.

I think the chip might be bad, and I'm kicking myself even harder for letting the other three go with the Fiero. I wasn't thinking much about the Camaro at the time, it was still 6 months away from being Historic.

The last thing I did just for the hell of it was pull out the chip from the ECM while the car was running during its hissy fitting. It made absolutely no difference at all, as if the chip wasn't there in the first place.

Does anyone have any spare AT29C256's or 27SF512's that they can overnight me? I have Paypal ready and waiting. It would be SUPER appreciated. Otherwise I'll have to see if Moates will do an overnight.