ZX-Tex

iTrader: (15)

FWIW here is another log from when I returned home and parked the car, engine still running, A/C on low, fans running continuously, voltage good. The plot shows everything that gets logged that is of consequence (no aux temperature, no VVT, no wastegate). Everything is smooth on the log, and the idle sounded fine with no popping, then it just died suddenly. You can see that everything on the log goes wrong simultaneously; there is not a preceding event that I can see.

Once things go wrong, looking at it time step by time step, the ECU ends up going to 100% DC on the idle valve. And the AFRs go high too. But, look at the injector duty cycle. It is rising fast, and eventually peaks at 9.5; that is 3-4X what it was at when everything was steady. So why is the AFR still going so lean?

Also, at about 136 seconds into the log, there are two very rapid fluctuations in the WT reading. I cannot think of anything that would make the temperature go from 91C to 85C then back to 91C in 0.5 seconds. It does it again right after that.

zipped csv file also included.

ZX-Tex

iTrader: (15)

Good catch. For whatever reason, I thought the adaptronic map was already set up for the GM open element IAT sensor; that is what I am using. But after just comparing to the xls data you posted in that other thread it is not. The temps were about what I would have expected when I looked so I had not double checked it or tried to calibrate it. I did calibrate the water temperature input.

I used your values in the table. I'll give it another shot and see if it helps. At any rate it needed to be corrected. Thanks!

Stein

iTrader: (46)

I'm guessing the values that you have for the IAT are what I supplied Travis. I had calibrated the IAT myself twice. I placed the IAT in a box with a thermometer and heated the air with a heat gun to 150* or so and "learned" the temp for each break in the table on the way down. The second time it was very close to the first time so I went with it. Mine is normally around 18*C, which has been ambient so I never questioned it. That said, the 15*C is really close so the margin of error is pretty good right around ambient. As I didn't have this table, I did it on my own. I'll update mine to reflect this as well.

In any case, I reset the fan back to 70* and warmed up the car. no change in RPM when it hit the fan, so the extra effort for electrical is working (I assume) as I would have anticipated a RPM drop otherwise.

I let the car warm up all the way to 105*, revving every 5*. Each time the idle settled down with no drama, no dip, good AFRs. I did this all the way to 105*.

I shut off the car for three minutes. I restarted and it was utter crap. It would not even stary running. It kep dieing, AFRs all over the place. Finally after a couple of starts I "helped" it catch with the throttle and it would idle, but AFR's were swinging. When I blipped the throttle it revved and then promptly died.

I did absolutely nothing to the car or the map. I just shut it off and restarted three minutes later.

I'll post logs tomorrow, as I don't have winZIP at home.

TravisR

iTrader: (9)

Just got off the phone with Rob, and he suggested switching to open loop, I don't know if anyone has tried that. He has confirmed that the brains of the ecu work way below 11 volts so there is absolutely no reason to think the ECU is failing. Regardless of what happens he reassured me we will find a solution to the problem, and we will get it fixed for you guys. Andy and Rob both have been working on this, and are keeping tabs on whats going on here. So your concerns are not falling on deaf ears, it just unfortunate that things turn slower then they need to be, especially when boost and throttle mashing are at the sacrifice.

What EGO are you running ZX?

Stein

iTrader: (46)

I did just that tonight and it MIGHT have been a bit better. First I changed the IAT tables as referenced above. Still no luck on a warm (not hot) restart.

I changed to open loop and it did start but was fluctuating a lot. At that point it may have just been due to the lower temp, as water was down to 42* by that time.

I'm going to leave it in open loop tomorrow.

BTW, in my hot restart, trim was pegged at 4 yet again.

ZX-Tex

iTrader: (15)

OK good, glad to hear they are still on the case. Thanks for talking to them Travis. If it is a tuning issue, I'll be happy to eat a huge plate of crow, as long as it gets fixed.

Do you mean switching to open loop for the idle, or for the AFR? I tried zeroing out the AFR closed loop PID values a few days ago but when I did that the Autotune stopped working; according to the manual, adaptive AFR tuning relies on the AFR closed loop correction factor.

If you mean defeating closed loop for the idle, that is easy. I can try that tomorrow.

For EGO I am using a Innovate LC-1 connected via serial. I also have an analog dash gauge. I did a free air calibration last week.

Stein I think it was your table IIRC. The new AIT calibration table seems to be giving me good air temperature numbers, at least based on a quick, low load drive around the neighborhood. Once I got air moving over the intercooler, driving about 30-40 mph, the AIT dropped to within a degree or two of ambient air temperatures, about 78F. But when I parked the car, it still eventually died. The fans were on the whole time, but the a/c was set low so that the compressor was cycling on and off. I would guess it cycled probably 15-20 times before it died.

FWIW the MAT reading had heat soaked to about 20F over ambient once I parked and idled the car. The GM IAT sensor is in the IC, in front of everything, but the condenser is only an inch away if that, and all of the ducting, compressor, etc. are just sitting there under the hot hood heating up what little airflow there is.

y8s

iTrader: (8)

just for fun, i took a log today with the car not running (it's good at that).

just key on and that's it. wasn't touching anything (no pedals, nothing)

this is what happens when you have a ground loop issue. and the odd thing is, it will behave normally for a while and then all of a sudden, do this crazy jazz. all the inputs go haywire.

the second graph illustrates this nicely. solution? I pulled the TPS ground out of the adaptronic-side connector. and all is well. 3rd graph (the spike is probably a key-off-on or something).

TravisR

iTrader: (9)

Tex,

We mean switching everything to open loop. What this is set to test is the trimming values. We want to make sure that the computer for whatever reason it is only trimming values when it should be. So set the idle to open, the fuel to open, and the ignition to open. The only thing you'll have now is built in trim for air temp and water temp. That we want to keep an eye on. You can also directly see what is being trimmed in the help and then calculations button. While the cars running it will show you the actual values.

ZX-Tex

iTrader: (15)

^^ OK will do Travis. I switched the fuel mode to open loop, ignition was already open loop, and zeroed out the closed loop idle settings (except for the target rpm map). I also zeroed out the closed loop PID settings in the closed loop fuel window just in case.

I ran a log to work this morning (no issues) and will run another one on the way home. When I get home I'll let it sit and idle while logging. It is going to be in the 90s today, and could hit 100F tomorrow, so that should provide some good heat soak test conditions

Holy crap on that ground loop noise Y8S...

ZX-Tex

iTrader: (15)

Guys, though I am tempering my enthusiasm, I might have just made a breakthrough...

So I turned off all of the closed loop stuff as mentioned above. At lunch, I decided to go for a drive and do some logging for Travis. I drove in some stop-and-go traffic, pulled into a Sonic, ordered my food, and let the car idle. Eventually it died, and I got it logged. Travis I was watching the fuel trim in the flow chart diagram under the help menu. I did not see anything unusual, trim was very low, like 0%-1%.

So I shut off the car, let it cool down, and ate my lunch. I noticed in the gauges window that the MAT was high, like 125 deg F. Heat soak, worse than I thought. Hmmm.... So I finished lunch, drove back, parked, and decided to try some tuning on the MAT air temp trim table. This adds trim to the open loop duty cycle based on MAT. This is when things started getting a lot better.

I had to go back and forth between the air temp trim table, and the AFR table. But basically, as the MAT would climb, the rpm would drop a bit, and I would add idle trim to bring it back up to about 1000 rpm. As the IAC valve let more air past, the mixture leaned out a bit (watching the AFRs), so I went to the AFR table and added a tad of injector time to the active cell, and maybe one of its neighbors depending on where the kPa was at. For example, if it was at 50 kpa, I added some fuel to the 40 and 60 cells.

By repeating the above process as the MAT rose higher, I got the idle to be very stable. I heat soaked the crap out of the motor, full A/C, coolant temps up to 220F, MAT up to 50C, and it DID NOT DIE! I cycled the a/c on and off, blipped the throttle, etc. and it still kept idling. Hoooooly ****...

I then took the car for a short spin doing some stop-and-go driving. No problems, even as the MAT and coolant temps went up and down. The car kept running well. I had one problem at a stop light, popping, rough idle, but it cleared up pretty fast; I was able to drive normally maybe 10 secs or so after I left the light. I think I needed to just apply a little more tuning, which I did after I parked. I let the car idle for awhile, doing some more idle trim/AFR tuning. Other than that, it ran well. It did not die. I had a little smile on my face.

So, attached is a zip with three items. One is the ecu file. The other two are the logs from before I did the tuning described above, and one that was during/after I did the tuning. Take a look at how I did the trim table in the open loop settings. Also take a look at the relative value of the cells in and around the 40 kPa/1000 rpm area on the AFR table.

Stein

iTrader: (46)

Sounds great! Mine was rough after lunch startup and kept getting better the whole way back to the office to substantiate your claim.

I'm sure that my "home calibrated" IAT table was a source of both of our consternation, as I kept seeing relatively flat IATs so kept runing out air temp heat soak and for that I apologize.

I took your ECU file and will be trying it out tonight! I did input the new IAT values as posted above last night.

ZX-Tex

iTrader: (15)

No need for apologies! Either way I am not sure that was the major part of the problem. I think the key is using the idle trim while hand-tuning the AFR to compensate. We may have been bitten by the ol' IAT heat soak issue I am not ready to declare victory yet, but things are looking up.

My injector times are going to be off from yours so you might want to try inputting the stuff by hand. You might also want to take a look at my calibrated settings for the stock coolant temp sensor and the MAP sensor. I think they are both reading pretty dang close to reality.

I am going to leave the closed loop stuff off and log the drive home. I'll report back on how it goes later on this evening.

Stein

iTrader: (46)

Yeah, I will be looking at it more for reference than anything. I know that I didn't calibrate the coolant and used supplied nubmers for the MAP. My MAP seems good, 106 at ambient, 42-43 at idle, 154-156 WOT so matching my 7 PSI wastegate pretty good.

y8s

iTrader: (8)

if you guys have a vac pump it's pretty easy to calibrate the map sensor almost perfectly.

suck it to whatever minimum you can reliably read on the pump and use that as the low end learn setting and hit the "learn" button.

then pull up weather underground (no relation to terrorists)
Omaha, Nebraska (68101) Conditions & Forecast : Weather Underground

and punch in your city and go to the "weather stations" link and find the closest few to you and use that value as the current map. then just start adjusting the upper value ADC (leave pressure at 300) and stop when it gets to the right atmospheric pressure value. incidentally, it only shows 3 sig figs (no decimals), but if it takes 5 ADC values to raise it from 100 to 101, you know each is about .2 kPa.

Stein

iTrader: (46)

My mightyvac guage isn't right any more. Something about hitting it with more than 30 PSI of positive pressure and pegging the guage. So, it's off a bit.

Travis looked up my local when we were on the phone initially setting mine up so I should be good on the top end.

ZX-Tex

iTrader: (15)

That is more or less what I did. I used a mighty vac, which agreed with the boost gauge. They even use the same scale, inHg. Once I did that for the low pressure ADC calibration, at -25 inHg IIRC, I checked the kPa value with no vacuum (atmospheric pressure) against the local barometric pressure at that time. It was right on, so I just kept the upper default value.

TurboRoach

iTrader: (3)

ZX,
It sounds like you adjusted the IAT fuel trim the same way I did. I got the car to idle when warm with low IAT (~15C) and then just let it heat soak and kept adjusting the IAT trim to keep the AFR I wanted. I'm pulling air off of the back of the rad so it didn't take too long to get it adjusted up to ~60C.

I'm using a Motorola MAP sensor and I calibrated it by watching the lowest value my gauge would hit on decel going down a steep hill. I put that in for low kPa and then did it again and hit learn. The high kPa I did the way y8s mentioned.

Hopefully fixing the IAT trim solves this for you guys.

ZX-Tex

iTrader: (15)

I forgot to mention... during this process, there were some times when the engine started idling very rough and acted like it was going to die. Same behavior as before. But as I added idle trim, and correspondingly corrected the AFR back to 13-14:1 or so, I could 'save' it and keep it running. That is what really makes me think this could be the solution.

So, to clarify what I did:
1. Start with a cool engine. Open the gauges panel so you can keep an eye on RPM and AFR. Have the air temp idle trim window (whatever it is called) open. Start with all set initially at zero. Or you can start with what I came up with, engine differences notwithstanding.
2. As the engine heats up, watch for an rpm drop, even a small one, like 100 rpm.
3. Look at the MAT reading on the gauge panel. Add trim to the corresponding cell in the trim table. Try adding 1 at a time. It should not take much. BE SURE TO HIT THE 'APPLY' BUTTON. If you are between cells, add to the closest one.
4. Once the idle returns to normal, add the resulting value to all of the higher temperature trim cells. So if you end up with '2' at 35C, add '2' to the 40C, 45C, 50C, etc. cells.
5. Check the AFR. If it is was rising much above 13:1 (no a/c) or 14:1 (with a/c compressor engaged), I added fuel to the corresponding fuel cells. Small adjustments, like +0.1, might do it if you are already close. You will have to close the trim window to get to the injector map. Be sure to hit 'apply' first.
6. Go back and tweak the trim cell if needed (due to the AFR change).
7. As the MAT heats up, repeat steps 2-6 for each of the cells in the trim table.

Buena suerte!

ZX-Tex

iTrader: (15)

OK good thing I did not declare victory. Not fixed... Died multiple times on the way home, in traffic, and in the driveway. I just love sitting in the turn lane, in traffic, with my hazards flashing, while my car stutters, backfires, and dies. LOVE that ****...

Here are tonight's fail plots. The first one is coasting to a stop into a turn lane after cruising at about 40-50 mph. I was finally able after a few minutes to limp it into a gas station, where I tried some more MAT idle correction tuning. The second one is parked in the driveway. It did not take long after parking for it to start running rough.

FWIW it seems to be a bit better but is far from fixed
Attached zip file contains logs and the ecu file, blah blah...

Stein

iTrader: (46)

I thought I was in good shape as well. Did the IAT idle tune, all was well. Started applying extra efort for AC, got increase on 100 RPM. But, for some reason, it won't take extra effort for electrical. I can't get it to run fan high, AC on and lights. It just surges.

So, I gave up on that after a while. FIgured I would go through the rest of the closed loop tuning. Fow whatever reason, tuning the PID per the link above doesn't seem like it is doing anything. I had P all the way up to 120 and idle was still stable.

Brought it down to what appeared to be a reasonable level, added in I and was idling OK. Touched throttle and it revved and held at 4K.

I don't know, I give up for the night. My wife won't appreciate 4K revs at this hour. Will try again tomorrow, likely starting over in case I missed something.