90 Beretta GT 3.1L rough idle

[DukenukemX]

Only has about 126k miles, but the engine has a rough idle. Even worse, the idle will bounce when stopping near a stop light. I've tried many things to correct this, but with little change.

So far I've replaced...
Idle Air Motor
Throttle body gasket
Plenum gaskets
EGR gaskets
wires
plugs
1 coil pack

I've also replace a good deal of vacuum lines, cause they literally broke from the touch of my hand. This piece in the picture was replaced with a T split. No idea if that was a smart move, cause I have no idea what it does. Few years ago I also replaced the intake manifold gaskets as well. There's no check engine light. I'm stumped as to what's going on. I've tried to take a can of carb cleaner and sprayed all over the engine to see if the idle changed. Nothing.



Uploaded with ImageShack.us

[Purple pit]

Looks to be a vacuum diaphragm that should be replaced.

[robertisaar]

no mention of an O2 sensor?

[DukenukemX]

The O2 was replaced years ago, about maybe 4-5 years ago? The item in the picture I figured to be a HVAC Heater Control Valve, or known as the air check valve. Part # is 14056648. Looks like this. Already ordered a replacement. So robertisaar you think I should replace the O2 sensor? I've had problems with replacing the O2 sensor on the Beretta. Don't remember what's in there now, but the Bosch O2 sensors never lasted.

[robertisaar]

Bosch don't seem to play well with anything GM, so not really surprising.

Denso, Delco, NGK. i wouldn't use anything other than those 3.

O2 sensor can do what you described, but you would need to monitor the datastream to be certain without throwing parts at it.

[DukenukemX]

Don't have an OBD1 reader. May have to look into getting one finally.

[3400-95-Modified]

ALDL cable and I'm sure someone can point you in the direction of free software that you can get scanning definitions for your ECU and get a real-time data stream.

[DukenukemX]

Already on it. Found a guide on how to make one. Ordered a Arduino FT232RL and a GM 12 Pin adapter which I'm just going to cut and make an adapter OBD1 to USB. Going to use TunerPro software to the best of my abilities. I'll be back in 2 weeks.

http://www.ebay.com/itm/260917151841?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

http://www.ebay.com/itm/390567697890?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

[robertisaar]

BUYER BEWARE!

i bought 10 of those cables from hohfashion, none of them are the cable that is pictured. it's actually a longer version, but it only has 4 wires actually connected, 3 of which are useful. you won't be able to connect to anything using a 160 baud datastream because it has no E pin. it does have A, B and M though.

if you really want one of those, i can get 1 to you within a few days since i got stuck with all 10 of them(since it would have cost as much to ship them back as what i payed for them).

[DukenukemX]

Took a look at my 12pin connector and I have A,B,M, and F. I just glanced at it, so I maybe wrong. F is for TCC? I'll wait to receive the connector, and if you're right I'll hit you up.

[DukenukemX]

Robertisaar turns out you're right, the cable only has 4 wires, but the two I needed are there and works. Also yea I'd say I'd be back in a month, took longer then I thought.

Anyway I got TunerPro RT working with it and recorded a data stream for about a minute. Car ran worse before I hooked up the computer, so I missed out on some good recorded data, but maybe someone here can figure out what's wrong? Also it now has a new Delco oxygen sensor, but it seems to make no difference.

[robertisaar]

O2 rich code was set through the entire log. i'm not sure if that is something left in the memory from the old O2 or what. clear codes(via mode10 commands or pulling the ECM fuse or disconnecting the battery). and relog.

BLM and INT are both indicating that fuel is being added on top of the base calibration, which would imply a lean condition. why lean is being detected by rich reported, i don't know.

you're idling at a somehow high TPS%, even though TPS volts are as expected. wonder if the TPS is getting intermittently bad?

IAC ran down to 0, indicating a vacuum leak or badly adjusted throttle-stop.

[DukenukemX]

O2 rich code was set through the entire log. i'm not sure if that is something left in the memory from the old O2 or what. clear codes(via mode10 commands or pulling the ECM fuse or disconnecting the battery). and relog.

BLM and INT are both indicating that fuel is being added on top of the base calibration, which would imply a lean condition. why lean is being detected by rich reported, i don't know.

you're idling at a somehow high TPS%, even though TPS volts are as expected. wonder if the TPS is getting intermittently bad?

IAC ran down to 0, indicating a vacuum leak or badly adjusted throttle-stop.

The throttle-stop is acting up. I've adjusted it, but when the engine runs the throttle is lifted higher then the stop, but not when the engine is off, and I have no idea why. Maybe the spring for the throttle body is weak? The TPS is pretty old and haven't changed it in years.

I'm going to disconnect the battery and change one of the coil packs. One of them went bad recently and I bought an extra. The one coil pack I haven't changed is 23 years old. I'm going to relog and post it here again.

[DukenukemX]

I changed the coil pack and unplugged the battery for at least 30 minutes. Started the car cold and I got a log of it running. It ran like crap for 2 minutes and then started to run better.

[robertisaar]

the throttle-stop NEEDS to be dealt with.... so many functions are dependant on throttle position being calculated correctly, any kind of intermittency here will cause all kinds of idle/low airflow issues.