Does it have to learn to idle?

[Sqrly]

My vehicle stalled this morning while warming up with the defroster on.

Skipping the diag, the RPM drops when the AC compressor turns off. I see the LTFT drop at the same time (and rise when the compressor turns on) The STFT fluctuates constantly but not a lot, this is normal?

I can manually turn the compressor on and off (snowflake button) and cause the rpm to fluctuate. It usually it drops and recovers.

It only recently cooled off here. I've been sitting idling with the AC running for 30 minutes a day with no issues.

Starting this week I'm running the defrost in the morning. It's overnight lows of 40+ and it's usually around 50 when I'm warming it up.

Notes:

Monday morning (10-28-24) I was greeted by a dead battery. Low enough the headlights would not turn off manually after unlocking the vehicle. It died because I shorted out the "it's down" switch in the midgate again when replacing the regulator.

I got a jump, drove it a few miles then fixed the switch when I got home and charged the battery. I did disconnect the battery when troubleshooting.

Now I am wondering if the battery dying and/or being disconnected made it forget how to idle.

 

[TJBaker57]

I have read that there is an idle learn procedure. I can remember reading it and thinking it sounded much like a normal startup but maybe with a shutdown added between two startups? Maybe someone else remembers?

 

[Sqrly]

I have read that there is an idle learn procedure. I can remember reading it and thinking it sounded much like a normal startup but maybe with a shutdown added between two startups? Maybe someone else remembers?

Since you put it that way .. (I searched for "gmt360 idle learn")

1) Start the engine and let it idle in park for 3 minutes. During that period the idle may be higher idle than normal.
2) After 3 minutes, turn the engine off and leave it off for 60- seconds.
3) Start the engine again and let it idle in park for 3 minutes one more time. During these run periods, you may see a check engine light. Scan for codes and if you get throttle body-related codes, clear them.
4) If the idle speed is still off, drive the vehicle at speeds above 44 mph and operate several acceleration and deceleration cycles. Then repeat step #3 one last time.


The idle should now be normal

Actually, I've done that organically. It only started the big rpm drops yesterday, 2 days after the battery disconnect.

Perhaps I'll disconnect it again and go through that procedure. Maybe it needs to learn without turning the compressor on.

 

[budwich]

I don't think a loss of power will cause a loss idle cells in the PCM. I think people usually only do a relearn if they have changed some "engine run" components such as a air sensor or cleaned the throttle body.... which is a likely starting point for "bad idle" especially if it hasn't been done in a while.

 

[Mooseman]

Classic dead/disconnected battery syndrome. If the battery went totally dead, then yes, you lost the idle settings and it went back to baseline with a clean throttle body. You'll have to pull the TB and clean it thoroughly. There is no other idle relearn to do. You should disconnect the battery again while cleaning it.

How to clean your throttle body

 

[Sqrly]

Classic dead/disconnected battery syndrome. If the battery went totally dead, then yes, you lost the idle settings and it went back to baseline with a clean throttle body. You'll have to pull the TB and clean it thoroughly. There is no other idle relearn to do. You should disconnect the battery again while cleaning it.

How to clean your throttle body

When I first got this vehicle in march of this year, the first issue I had to deal with was the parasitic battery drain which turned out to be the instrument cluster.

The battery was disconnected many times and I never did an idle relearn. Won't it sort itself out in a week or so?

 

[Mooseman]

No because it`s pre-programmed to adjust the idle as the TB gets dirty. It now thinks you have a clean TB and tries to idle but it can't because the TB is actually dirty. You have to clean it.

 

[59840Surfer]

Just a heads-up thingy ----

I had a couple of dead battery cars in here in the past few months --- and they all had one commonality --- OnStar.

OnStar is dead --- zilch --- and the radio in the vehicle is trying to talk to the Mother Ship --- which is gone now.

It's possible (and likely) that IF OnStar is calling home 24/7, then it can kill a marginal battery overnight and a new-ish battery in a couple of days.

............... just sayin' ...................

 

[Sqrly]

✓ Clean TB
✓ Relearn Idle
✓ Happy ECM

Note: There was no need to remove anything but fuse #10, the resonator, and the TB. I did unplug the intake air temp sensor to move the wire out of the way.

Now, does anyone know how to clean out the intake manifold without removing it?

 

[Sqrly]

Just a heads-up thingy ----

I had a couple of dead battery cars in here in the past few months --- and they all had one commonality --- OnStar.

OnStar is dead --- zilch --- and the radio in the vehicle is trying to talk to the Mother Ship --- which is gone now.

It's possible (and likely) that IF OnStar is calling home 24/7, then it can kill a marginal battery overnight and a new-ish battery in a couple of days.

............... just sayin' ...................

Actually, I've never given Onstar a second thought, or a first for that matter. My vehicle has the buttons on the mirror and I believe the box is behind the glove compartment.

I only need to pull fuse #27 under the back seat to disable it correct? Why have it powered up if it can't work.

 

[Mooseman]

The OnStar module is located under the RR seat. I just unplugged it. Sometimes fuses power other things that we don't know about.

 

[TJBaker57]

Mine is still connected.

Cannot say if it draws any considerable power as I keep a battery tender on the vehicle due to very low vehicle usage.

On a related note: did you all know the Tech II queries both the radio and Onstar before it does anything else? I have no idea what the data is but it requests a single memory block from each of them.

 

[Sqrly]

The OnStar module is located under the RR seat. I just unplugged it. Sometimes fuses power other things that we don't know about.

Ah ok. I was thinking of the wrong thing, the one behind the glove box is the XM yes?

I just pulled the entire On-Star box out, took about 2 minutes, no tools required. I replaced it with 2 foam blocks and sandwiched the wires between them to keep them from rattling around.

After, I checked a few things. Garage door opener, compass, radio, and I even checked for trouble codes. Everything else still works as it should.

 

[Mooseman]

Mine is still connected.

It's useless now as there are no analog cell signals available for it. It might draw a little more as it continuously searches for something that no longer exists. Might as well disconnect it.

Ah ok. I was thinking of the wrong thing, the one behind the glove box is the XM yes?

IIRC, yes it is however I never had one .

 

[Sqrly]

I've already cleaned the throttle body. (see post here) It idled fine for a few days now it stalls again when the compressor shuts off.

Not when it engages, when it disengages. I watched the clutch to confirm this.

It idles fine otherwise. It idles high when first started when it's cold ( 38f ) and it warms up in just a few minutes. I'm inclined to suspect it is related to the outside temperature and not the low battey / failed to start I had about the same time it started. (Which was caused by me shorting out the mid gate window switch, again)

The overnight lows here have dropped 15 degrees from Oct to Nov. This issue didn't appear until it got cold enough that I started needing to defrost the windows at 7:30 in the morning. I never had a problem running the AC while idling in the summer (90-100 degrees) for 30 minutes at a time even after several dead battery issues and never cleaning the TB.

I have the Car Scanner ELM OBD2 app on my android and @TJBaker57 has helped me both learn how to use it and make it more useful.

Below is a capture of when it almost stalled with short term fuel trim, long term fuel trim and rpm. The next time the clutch disengaged it did stall. When the clutch engages it's very smooth with only a small momentary increase in rpm. The STFT never smooths out, it's always fluctuating like seen is the first half of the image. It only stalls once warmed up, not while in the the cold / high idle state.

 

[Mooseman]

Those trims are when it's warm? The STFT seems a bit on the rich side as it's taking fuel away (-10). It's normal for it to be a little bouncy but at -8 to -12, it seems to be running rich and it's trying to adjust more than usual. LTFT also tries a lot to pull fuel just as it stalls. Then to recover, the STFT is dumping a bunch of fuel (+20).

There are two fuses that supply the PCM. I'm not sure which one that actually resets the PCM but try pulling them both for 30 minutes. Even though some say that just one fuse is required, I pull both just in case.

 

[budwich]

I assume the time scale is in minutes. As the "moose" says, it would appear that the engine is running rich in terms of O2 feedback to the PCM. The PCM is trying to lean out things by negatively moving the trims. The sum of the trims is pretty negative considering the system is intended to "normalize" around 0. Reaching a sum of 25% for a period would throw a code ( lean or rich accordingly to the signed value).

Something very weird. It is somewhat too bad that you didn't have the O2 voltage at the same time. Not sure what the reading would be but it probably was either very high or low for a "long" period... again this probably should have thrown a code.

As asked, what was the engine coolant temp?

 

[Sqrly]

There are two fuses that supply the PCM. I'm not sure which one that actually resets the PCM but try pulling them both for 30 minutes. Even though some say that just one fuse is required, I pull both just in case.

I only pulled #10 when I cleaned the TB. I was thinking something similar, going ahead and disconnecting the battery for 30 min to reset it for sure.

Something very weird. It is somewhat too bad that you didn't have the O2 voltage at the same time. Not sure what the reading would be but it probably was either very high or low for a "long" period... again this probably should have thrown a code.

As asked, what was the engine coolant temp?

I believe I could log the O2 voltages. The dash gauge said the temp was at 210.

I did some more testing while waiting in front of the school today. When it is 61f outside the issue is the same.

FYI It idles smooth as silk when it's in drive and not moving. It's only stalling when it is idling in park.

This new capture shows when the ac clutch turned off then back on while ilding in drive with the parking brake on. My foot was not on the brake pedal.

Note: Could it just be bad gas? My regular (Chevron) station was closed for repairs for 3 tanks of driving, and I was using an Arco that just went up nearby. My regular stop reopened and I just filled up with my regular Chevron (with techron) yesterday.

(Going to pull it in and disconnect the bat now)

 

[Sqrly]

I disconnected the battery for 45 minutes.

I reconnected the battery and turned the key to ignition and waited until all the various noises stopped.

I started the engine and let it idle for 3 minutes. (Did not touch the pedals)

I turned it off for a minute. (I opened and closed the door to let the RAP turn off too)

I started the engine again and let it idle for 3 minutes. (Did not touch the pedals)

I shut it off and let it sit for about 5 minutes.

I then drove it around accellerating to 65, sometimes softly, sometimes briskly, stopping several times. (We have streets and avenues on a 1/2 mile grid out here in farm country) One time manually downshifting to 3rd and then 2nd as I came to a stop.

I'll post again after I warm it up tomorrow morning as I do every cold school day. (Let it idle with the defrosters on while I squeegee the windows)

 

[Sqrly]

I assume the time scale is in minutes. As the "moose" says, it would appear that the engine is running rich in terms of O2 feedback to the PCM. The PCM is trying to lean out things by negatively moving the trims. The sum of the trims is pretty negative considering the system is intended to "normalize" around 0. Reaching a sum of 25% for a period would throw a code ( lean or rich accordingly to the signed value).

Something very weird. It is somewhat too bad that you didn't have the O2 voltage at the same time. Not sure what the reading would be but it probably was either very high or low for a "long" period... again this probably should have thrown a code.

As asked, what was the engine coolant temp?

During the second 3 minutes of idling mentioned in post #19 above I monitored the O2 Voltage. The capture is only 15 seconds but it stayed like that the whole time.

 

[budwich]

The O2 looks "as expected"... waggling along and it appears to follow the STFT... ie. leaning the fuel resulting in lower voltage (more O2). The problem is the value of the STFT... they are significant leaning. Is it possible that you have setup the graphing values in some form. I seem to recall that you provide ranges on the graphs x / y.... but I haven't used the graph feature in a while. I just monitor realtime values with gages.

IF things are as shown, AND the LTFT goes significantly low, then it is quite likely the engine will be "fuel starved" at instances during those periods as the system won't be able to "catch" the system with STFT fast enough all times as seen in the first picture post probably resulting in a stall / stumble depending.

 

[Mooseman]

Might still be early. Give it a full day of driving and see if it stabilizes. That is close to the 25% threshold. Check for codes to see if one might be pending.

 

[gmcman]

The STFT will try to balance out the LTFT as much as it can to 0 or within 0- about +-7 (don't quote me on the 7). At idle the STFT will vary much slower than with the engine at a higher RPM while driving. The LTFT will sometimes also show high values when decelerating. A slow changing STFT while driving generally depicts a lazy upstream O2 sensor.

How many miles, and how old are the plugs?

What are your volts when driving/idling? Reason I ask is you said you had a short and want to be sure your diodes aren't messed up.

Check your glovebox light for being warm as soon as you open it, would be a sign of it staying on.

 

[Sqrly]

It did not stall this morning after fully warmed up.

Perhaps only pulling fuse #10 is inadequate? Throttle body (TB) relearn (4.2)

I also did not do the 'accelerate and decelerate past 44mph several times' on the first try.

The LTFT looks a lot different now (The STFT looks the same). I see many more smaller negative steps now. I only 'hear' the rpm go up a little when trim goes up. Before I heard a huge drop when the trim dropped.

I checked the coolant temp with the app. It says 190 when the dash gauge says 210.

It has about 190k on it and I changed the plugs when I got it, putting in the proper ac delco ones.

Note: The captures I usually put in my posts are screenshots taken in real time not from log files.

Edit: Now that @TJBaker57 told me I can zoom the log files, here's the current trim cycles.

 

[TJBaker57]

Note: The captures I put in my posts are screenshots taken in real time not from log files.

I'm adding 3 sequentail captures to show the entire LTFT cycle. If I can control the scale, I do not know how. It automatically resizes.

If you have enabled data recording then everything you display while driving is available in a logfile.

In the logfile you can select what parameters to display, whether to display on one graph or multiples, and you can zoom in or out of the graphs. The zoom can be either one of the two axes or both simultaneously.

As far as I know you cannot dictate what the X and Y axes display, they will automatically adapt to the level of zoom.

 

[TJBaker57]

For a stalling issue I would be curious about what the throttle position, A/C compressor feedback status, injector 1 pulse width and such are doing during these events.

Like does the throttle close significantly when the compressor kicks off? Is the throttle opening in response to an rpm drop?
What are the injector pulse widths doing?? Are they giving more or less fuel?

 

[Sqrly]

What are your volts when driving/idling? Reason I ask is you said you had a short and want to be sure your diodes aren't messed up.

I shorted a terminal of the midgate window postion switch to ground after modifying it to remount it with the tab broken off. Then I did it again when replacing the regulator a second time because the first one failed under warranty. I've redone it to eliminate the issue now.

SOLVED! Impossible midgate switch problem

 

[Sqrly]

In the logfile you can select what parameters to display, whether to display on one graph or multiples, and you can zoom in or out of the graphs. The zoom can be either one of the two axes or both simultaneously.

Cool, I did not know that. I changed the pic in my post above.

For a stalling issue I would be curious about what the throttle position, A/C compressor feedback status, injector 1 pulse width and such are doing during these events.

Like does the throttle close significantly when the compressor kicks off? Is the throttle opening in response to an rpm drop?
What are the injector pulse widths doing?? Are they giving more or less fuel?

It's not stalling currently after I did the TB relearn again.

I looked through the logs, the throttle position info is intermittant, mostly missing, from the logs. Perhaps if I actually monitor it in real time it will get higher priority in the app and record the data more reliably.

 

[Mooseman]

It's still pulling a lot of fuel on the STFT. How's the MAP sensor? New or original? Do you have a MAF? Might need a cleaning. How's the air filter? Maybe the throttle body is having a hard time regulating air at idle. What's the fuel trims while driving? At the age it's at, the TB's gears might be getting worn but not enough to trigger a fault code. Just thinking aloud here.

Beyond those, we're looking into the fuel system for a rich condition.

 

[TJBaker57]

looked through the logs, the throttle position info is intermittant, mostly missing, from the logs. Perhaps if I actually monitor it in real time it will get higher priority in the app and record the data more reliably.

The Car Scanner logfile records ONLY the sensors you have selected for monitoring during the drive.

During the drive/recording, if you have sensors displayed on a single graph, and you tap the right spot on a sensor in the legend, this turns off that sensors' display in the graph AND in the recording as well.

 

[Sqrly]

It's still pulling a lot of fuel on the STFT. How's the MAP sensor? New or original? Do you have a MAF? Might need a cleaning. How's the air filter? Maybe the throttle body is having a hard time regulating air at idle. What's the fuel trims while driving? At the age it's at, the TB's gears might be getting worn but not enough to trigger a fault code. Just thinking aloud here.

Beyond those, we're looking into the fuel system for a rich condition.

MAP = I do not know. It's there, it's plugged in. If it never failed I doubt it was ever replaced.
MAF= No
Air filter = I put in a new stock one when I got the vehicle in march of this year. It had a K&N that had never been cleaned. Apparently the previous owner thought 'lifetime' meant it lasted forever with no maintenance.

Fuel trims while driving? Define "Driving" and I will run a log to post.

This vehicle is mostly used to take kids to schools that are less then 2 miles away so it usually doesn't get over 35 mph. It see's freeway speeds maybe 5-6 times a month and even then usually only for a few miles.

 

[Mooseman]

During your drive to school at steady speeds. Would be good to see what's happening while not at idle. Wondering if it could be a fuel.control.problem with too high pressure that the regulator is defective during low fuel demand. Check the fuel pressure regulator for fuel in the hose. It can fail there and allow fuel into the intake.

 

[budwich]

MAP = I do not know. It's there, it's plugged in. If it never failed I doubt it was ever replaced.

I would suggest that you replace it... its relatively cheap and based on my experience can cause the type of thing that you are seeing with "wacky trims".

 

[Sqrly]

During your drive to school at steady speeds. Would be good to see what's happening while not at idle. Wondering if it could be a fuel.control.problem with too high pressure that the regulator is defective during low fuel demand. Check the fuel pressure regulator for fuel in the hose. It can fail there and allow fuel into the intake.

@Mooseman Do you have the CarScanner app to import log files into? "2024-11-15 07-42-06.brc"

I managed to copy the log from my phone to my tablet so I could 'literally' get a bigger picture. Most of the drive is spent accel/decel or coasting in drive. There's a few small spots of cruising at a fixed speed.

 

[Sqrly]

I would suggest that you replace it... its relatively cheap and based on my experience can cause the type of thing that you are seeing with "wacky trims".

To me $65 is not "cheap". If I need it I will buy it but I don't want to just throw parts at it.

There are cheap $10 versions but I use at least OEM quality if not actual OEM parts.

 

[Mooseman]

Had a look of your log file on my phone and one thing I noticed is your MAP is weird. There's just a single spike towards the end of your drive and that's it. No other data from it. Just for the heck of it, try unplugging it and running without it for a drive. Log the data. It should use default values according to throttle data.

I hadn't seen the cost of the MAP in a while either. They sure went up!

 

[Sqrly]

Had a look of your log file on my phone and one thing I noticed is your MAP is weird. There's just a single spike towards the end of your drive and that's it. No other data from it. Just for the heck of it, try unplugging it and running without it for a drive. Log the data. It should use default values according to throttle data.

I hadn't seen the cost of the MAP in a while either. They sure went up!

Since you made the same mistake I did, It must be a bug in the app. See post #30

I was only monitoring 4 things and yet when I go back to look at the logfile I made I see switches to view many more that it apparently didn't actually record data for.

I got a screenshot of that from a previous log I mentioned in post #28. I was only monitoring STFT, LTFT. and RPM. If I try to view data from any of those things that I wasn't actually monitoring, I just see a spike or 2 like in your photo.

I will run a new log to monitor anything you guys suggest. I really appreciate all the help.

 

[TJBaker57]

noticed is your MAP is weird. There's just a single spike towards the end of your drive and that's it

I would not judge any MAP irregularities by this logfile.

When viewing ALL sensors itnis seen that the only MAP data conicides with around ten or so more graphs at the beginning of the file.

This suggests that perhaps the MAP data was not active in the app during the recorded drive.

 

[Sqrly]

I would not judge any MAP irregularities by this logfile.

When viewing ALL sensors itnis seen that the only MAP data conicides with around ten or so more graphs at the beginning of the file.

This suggests that perhaps the MAP data was not active in the app during the recorded drive.

We were typing at the same time

 

[TJBaker57]

was only monitoring 4 things and yet when I go back to look at the logfile I made I see switches to view many more that it apparently didn't actually record data for.

I find the best logfile results are made when during the drive you choose to display the selected data points on ONE graph as opposed to separate graphs.

This ensures you get the data points you select in the logfile.

While I do sometimes mess with the display during a drive, like disabling the legend display or some such I highly suggest concentrating on driving and not the phone.