90 Beretta GT 3.1L rough idle

[DukenukemX]

Now that I look at the log, there's still too much air going into the engine. But I still can't find that leak. Is there a sure fire way to find this leak? I'll probably remove the plenum tomorrow to hunt down this problem. The PCV is letting a lot of air in, cause the suction is so strong that the line that comes from the front valve cover is strong enough to hold a credit card. Also the hissing is very audible when I remove the oil cap. Not sure if that's normal. For the hell of it, I got another PCV from AdvanceAuto with no change.

[DukenukemX]

are you saying it would idle at 2000 before with no DTC? if so, that would be due to TPS% being calculated as higher than 0%, so the DTC test would have been disabled.

I had a $6 TPS on the engine that would cause a 2000 RPM idle. I put back an older TPS that I had lying around. Older TPS might have been bad, but was never sure. The car had an issue with a bad lock cylinder that disabled it for a while, until I got a replacement. After starting it up with the new lock cylinder, it ran at 2k rpm. I didn't let it run long enough to see if a DTC came on. My instinct was to try the older TPS I had lying around, and the idle dropped. I swapped the sensors back and forth to make sure.

[DukenukemX]

I have a log of the engine running with me playing with the stopper. The engine was warmed up, and battery unplugged just before this. Near the end of the log I began to lower the engine idle with the stopper and the check engine light would turn off. But mostly not until I flicked the throttle up by hand, that the idle would drop and the engine light would go off.

To me it sounds like TPS is bad, which I already have a new one coming along with new wire harness for it.

[robertisaar]

DTC35 sets when there is 0% TPS and idle speed stays either 150 or 300RPM above the target idle for 30 seconds. or something like that, i'm going off of memory at the moment.

are you saying it would idle at 2000 before with no DTC? if so, that would be due to TPS% being calculated as higher than 0%, so the DTC test would have been disabled.

[DukenukemX]

Ok I haven't stopped trying to fix this problem. I just recently replaced the IAC and now get a check engine light with '35 IAC High' being displayed. The engine seems to idle fine and then suddenly this light appears and the engine idles a 1200 - 1300 RPM. Though before this the idle would be able 2k until I put my old TPS back in the car. The new $6 TPS from Ebay already failed it seems, though I expected it fail at that price. So I ordered a new Standard TPS and it's on it's way.

The IAC also produces '35 IAC High' error, so yea.

[robertisaar]

at full temp in park, you should be sitting around 700, though i don't know the exact calibration you're running. you can watch desired idle speed in the datastream to see how close it is to being at controlled speed.

you can close the throttle-stop while running without problems, since the ECM will relearn the minimum TPS voltage every time it goes down, but opening it while running will cause TPS% to start rising and cause issues.

[DukenukemX]

still not seeing any IAC counts at idle. not sure if air will leak through the TPS mating surface or not. i assume the throttle shaft bushings would be fairly tight to begin with and not allow much, if any air through.

Not the shaft but the bolt holes. When I removed the sensor I could see straight inside somewhere through the holes. I stripped the bolts probably a decade ago to change the TPS. Must have forgotten about it. Not saying that's the air source, but I took the extra precaution and put new bolts.

BLM looks to have shot down to 100(so quite some fuel was being removed to keep it stoich), so airflow is being overcalculated. this could be due to the TPS% value not being calculated correctly or the throttle-stop position.

what i would do(and i wouldn't say you're obligated to do it, just based on my experience).... is if i were stuck with a stock calibration, is if all other possible vacuum leaks had been fixed(and they look like they might), would be to adjust the throttle-stop and clear the ECM's memory to force a relearn until at least some IAC counts at idle were required to hit the target idle speed at full temp(5 or so would be enough for me). at that point, i would only modify the throttle-stop position further to get idle fueling closer to optimal.

That's what I'll do then. I guess I'll keep my laptop connected and watch for a BLM of 128 to be optimal. Though when the engine did finish the idle relearn after this post, the engine ran exactly at 1000 rpm, give or take 50. I'll check the BLM tomorrow to see where the engine sits in BLM before deciding to clear the ECM and do another relearn.

[robertisaar]

with a closed throttle, TPS% was being reported as 1.57 with a normal looking closed voltage of .53 volts. this can cause a high idle and fueling quirks. may need to clear the ECM's memory to let it adjust to the new TPS.

still not seeing any IAC counts at idle. not sure if air will leak through the TPS mating surface or not. i assume the throttle shaft bushings would be fairly tight to begin with and not allow much, if any air through.

BLM looks to have shot down to 100(so quite some fuel was being removed to keep it stoich), so airflow is being overcalculated. this could be due to the TPS% value not being calculated correctly or the throttle-stop position.

what i would do(and i wouldn't say you're obligated to do it, just based on my experience).... is if i were stuck with a stock calibration, is if all other possible vacuum leaks had been fixed(and they look like they might), would be to adjust the throttle-stop and clear the ECM's memory to force a relearn until at least some IAC counts at idle were required to hit the target idle speed at full temp(5 or so would be enough for me). at that point, i would only modify the throttle-stop position further to get idle fueling closer to optimal.

the problem you're having is largely the way idle fueling is calculated in a stock calibration. it isn't done in a speed-density fashion, rather more like alpha-n(or perhaps even MAF) with SOME limited input from the MAP sensor and calulated barometric. almost the entirely of the airflow calc at idle is determined by a combination of TPS% and IAC position. why GM chose this method, i have no clue. it's one of the first things i disable in any calibration i play with. if a TPS or IAC DTC gets set, the alpha-n idle fueling calc is bypassed and it is done in speed-density.

[DukenukemX]

I got a new TPS I ordered from Ebay for $6, and decided to put it in. I have bad experience when it comes to buying cheap sensors off Ebay, but for $6 I figure why not. I haven't changed the sensor for a long time, and noticed the threads to the bolts were basically gone. The sensor was held in, but hardly any threads. I figure maybe that's where the air leak is coming from? Put new bolts with the new sensor and started the car. Immediately I noticed it ran better.

Did an idle relearn and drove it for a day and the engine ran very smooth. Only problem was high idle. I remember buying a new IAC for the throttle, but again cheap Ebay Chinese made junk. I put it in 5-6 months ago, and noticed the engine idled higher. Still ran crappy, so I didn't pay much attention. I kept the old IAC, and decided to put it in. TunerPro was reporting too much air again, so that's when I put the old IAC to see if it dropped the air flow. It seems to have work, and the engine ran too low in idle. I had to adjust the stopped to increase it.

Please check out the log and see if everything is fine now. I'm doing an idle relearn, cause the idle does bounce a little with the new IAC, but it did for the TPS as well until I did the relearn.

[3400-95-Modified]

Yea it turns out it's the "thing" next to the battery. I'm guessing it controls the cruise control in the car. It's on a bracket, and it closes the gap on the stopper if I push against the bracket. I guess at some point the bracket was leaned on by someone and it would hold the throttle slightly open. I figured this out when I tried to find a vacuum leak today.

I started to wonder about the IAT sensor, the one sits on the intake. I've never changed this sensor and wonder if that's effecting the reading of high air flow. Cause I've been all over the engine and can't find a leak.

Try adjusting the cruise control cable to see if you can release the preload on it... it should have some adjustment to it.

[DukenukemX]

Oh yea I did remove a vacuum hose from the TB to check for fuel.

[robertisaar]

if you had the vacuum clip disconnected from the TB for about 2 seconds or more while running, that can cause the MAP code to set. or any loss of vacuum to the sensor, really.

[1988GTU]

I'd redo the unhook battery and let it relearn again. Then take a reading.

[DukenukemX]

I wonder if it's because I threw some carb cleaner at the MAP to check for a leak?

[robertisaar]

you're VERY close now.... still idling with no IAC steps, but advance is finally coming up and you're getting closer to target idle speed.

airflow is calculated airflow in grams/sec, not the air/fuel ratio.

fueling is VERY close as well, about as good as you'll get with a factory calibration anyways.

not sure why the MAP high DTC got set though... signal looks fine, i don't know if the ECM is using an estimated/calculated MAP value or not though.