Beretta.net

I am starting to think the suggestion on the ECT sensor may be a good direction on this -maybe once it warms up a little (is 30-60 seconds actually enough though?), a bad sensor could start giving a readable signal that isn't confusing the ECM. However, a faulty O2 sensor circuit might be causing this issue, and 30-60 seconds in the exhaust stream would definitely warm it up a bit... If the ECM is being tricked on the fuel mix, it could be messing things up enough to run that high idle. I really think you need to see the data-stream from the ECM, even confirming whether or not the ECM is in open or closed loop might point you to the problem. The rest is just shooting in the dark here...

I checked for vacuum leaks, cant find any.

Is it possible the computer is failing?

Just in case: I had a good ground on the TPS and a good 5v ref. My problem was the signal wire that goes to the ECM. Also it would show up every now and then. I kept thinking it was bad quality TPS sensors.
Did you clean the MAF?

Sorry -the intake boots are really problematic on these cars -are you CERTAIN the boot isn't cracked somewhere? Around the plastic tube from the valve-cover, or the IAT sensor maybe? How have you been checking for vacuum leaks?

I am wondering if some carbon dislodged into the new IAC valve maybe... The other possibility could be a bad EGR valve, but it would also need some kind of cracked tube to allow air in, or the idle would DROP, not rise.

Edit: What was the ECT reading prior to start-up?

A bad ECM is always possible, but as you say, very unlikely. Pulling the ECM connectors off and checking all the pins for corrosion or other damage would be a good step though, if not already done.

I don't recall that the OBD2 requires an idle re-learn, please correct me if I am wrong. You really have us stumped with this one!

Did you check the hose to the brake booster as part of the vacuum check?
When it idles high, pinch off that hose with some needle nose pliers and see if idle drops.
I know my booster used to lose vacuum in the cold winter weather, maybe a leak is there.

And just a shot in the dark, unplug your spark knock sensor and see if anything changes.
I have no theory about this, just that I had a slight stumble I chased by changing all
sensors and it turned out to be a bad knock sensor.

"The signal wire going to the ECM" <--good idea. usually found by wiggle checking. It seems like your car's problem here is too faint to respond to this as its one restart away from cure. As our cars age, any wire insullation exposure to the elements sets up corrosion and will start intermittant problems. We can rise to the occasion and we are ready for it. If the car were in my driveway, at any point of the day, I would check continuity of this wire by means of a long jumper wire from the engine bay [connector] to a small paperclip out of the corresponding ecm connector pin and alligator clip. I like needle meters for this test. watching the meter, wiggle check that entire harness. I wouldn't bother removing the glovebox for wiggle checking because a wire fault is most likely in the bad environment, the engine bay. If the computer connectors are corroded, then the wires out of them may be too.
-----------
I've not used the big $50k scan tool to monitor all systems. Man that would be cool. Also as a result your result figures are above my head.

I've always isolated systems one at a time.

I have never accepted defeat. We will win.

Your results do seem normal to what I think you were getting at.

I wonder what happens if we fooled the temp sensors [intake manifold and intake tube]. If the temp is sending warmer than the car actually is, we may learn more. I want to know what part of this car needs to be warm in order for it to behave.

A hairdryer works well for this! I'd want two of them, put one on the temp sensor, one on the intake sensor and leave it for 3min to bring that temp up. [try to keep the tps and Iac cold by rags to block heat]

If we learn that the car doesn't care, and the problem repeats with signal sending something warm like 40F deg upon the first start, then the problem lies elsewhere, like that IAC or TPS system.

If the car does care, then we can focus on the temperature system.

I realize this risks a false positive in that we have one shot to make the car run at first start when warming spring temps may cause a normal start anyway.

--------------------

my experience with bad computers comes with the SES light not triggering at proper time [like initial key on = light on not happening] and performing a wiggle check on the board while the car runs. Thats right, board removed from its case, and twist lightly holding the board on each side. I don't think this is us - yet. i'm not beyond taking the hair dryer to the ecm yet, though I'd be careful as unequal heating is reason for failure over the years. I'd probably opt to remove the box and bring it inside. Have breakfast with it, and both go out in the morning and attack the problem together

--------------------

Lets find what system is unhappy -15F deg

Swapped the computer, it was bad, car works again.

Nice to hear that you seem to have found the problem!

D@^$!! it! A bad ECM is the car just cheating! Really glad you seem to have found it, really surprised it was the computer. Nice that you had a spare!

Yep, verified.
It was the ECM.
Car has been running perfectly fine.
Thanks everyone for the help!

Good to hear it's fixed, I bet your pleased!

ifixalot wrote:Good to hear it's fixed, I bet your pleased!

Thanks, I am very happy it is fixed.

That is a real good example of multiple things wrong tripping all over each other confusing us! What a day!