Chronic Stalling Problem 93 Beretta

[3X00-Modified]

Yeah, get the pump replaced... It may solve most of your issues... Then work on getting a good reading on that coolant sensor and see if it is truly saying its -30... if so you need to figure out which sensor it is and get it replaced since that will really affect the running of the engine.

[Wisconsinite]

I'm going to replace the pump myself. Seems like it could save me a few hundred over taking it to the shop. Just need to pick one to buy..

I cleared the codes and limped the car into town. It had 8 codes from the 12 mile drive.

42 EST (spark timing) circuit
33 MAP high voltage, low vac -already replaced
32 EGR
24 VSS (vehicle speed sensor)
23 IATS high temp -already replaced
21 TPS voltage high -already replaced
14 coolant high temp
13 02 open circuit -already replaced

I think some of these are caused by the low fuel pressure. I'm wondering if a vacuum leak could be causing the EGR and MAP codes. I wondered if the computer could be on the fritz, but I chatted with a mechanic at the Chevy dealership and he said my ECU is probably fine, and that it is throwing codes on good sensors because of the havoc going on under the hood.

[Money pit Beretta]

The computer is programed to keep the car running if it is at all possible. Let me put it this way: My 90GT's ignition lock bars broke and the key was stuck on. I pulled the three front plug wires and it kept running. I killed it with some light power braking. Only just after the panic went away did I think about the injector fuse. I never thought that the computer could keep an engine running with only one bank firing. Must have been working overtime doing that.

[Asylum]

ECU failures are really pretty rare. GM actually did a good job on some things, LOL!

Replacing the pump yourself is not all that intimidating if you have a place to do it, and get most of the gas out of the tank.

Straps if needed I believe are available from Rock Auto.

Which pump you decide is up to you, your budget and your future plans.

I ran mid-13's on a stock pump with my 3500 FWIW!!

Cheers

[Rettax3]

ECU failures are really pretty rare. GM actually did a good job on some things, LOL!
Replacing the pump yourself is not all that intimidating if you have a place to do it, and get most of the gas out of the tank.
Straps if needed I believe are available from Rock Auto.
Which pump you decide is up to you, your budget and your future plans.
I ran mid-13's on a stock pump with my 3500 FWIW!!
Cheers

...And the stock pump in my GTU ran the 3800 SC just fine too for quite a while, before it finally succumbed to old age at 160K+ miles and over twenty years of age (at least I am pretty sure it was the original pump...). The most important part of doing the tank swap, aside from having it as empty as possible, is getting the car up high enough.

Even the ramps I had here didn't give me the room I wanted to drop the tank, so I used my long-frame jack for more clearance.

Also watch yourself around the filler-neck, that can be a PITA, otherwise, the fuel pump is a piece of cake to swap.
It seems that some people are always trying to blame the computer, and Asylum is right, they are (figuratively) almost bullet-proof. I am glad you seem to know better and are getting some decent advice. That said, they can fail, I have seen it happen a couple of times. It is just really unlikely.

[Cliff8928]

ECU failures are really pretty rare. GM actually did a good job on some things, LOL!

Yeah, I had the one fail on my GT. It made driving interesting when I'd have it die out with any throttle. Reach over and tap the ECU, then it would work again for a while.

[2.2H8TA]

Take a look at the filler neck too, when you're doing all this. It's worth it to replace, obviously if it's bad.

[Wisconsinite]

Bosch pump assembly arrived today.

The tank came out fairly easily, and the pump is a perfect match. But I managed to lose the wiring harness that attaches to the pump.

I've torn the garage apart, looked everywhere. Can't find it. Hopefully I can find it tomorrow morning. It'd be a shame to have to buy another one or rig something up.

[Wisconsinite]

You all still out there?

I got the tank mounted back on the car. I couldnt find a wiring harness anywhere. The closest I could buy is the plastic 5-pin terminal that plugs into the pump with 3 inches of wire.

Undaunted, I soldered on more wire, intending to match the wires on the pump end with those on the car. When I got the pump wired up to the car I tried to start it. The fuel gauge goes bonkers and the pump doesn't run. Clearly I have made a wiring mistake.

Inside the old pump I see a grey wire that connects to the pump mechanism, a black wire that connects to the pump and the gauge, and a purple wire that only connects to the gauge.

I am pretty sure the black is negative, and grey is positive. I applied 12 volts from some AA batteries in series.

Positive charge applied to grey and negative to black = pump runs
Negative to grey and positive to black = pump runs, but poorly

On the car there is a 3-pin terminal with the locking clip on the left. I'm pretty sure the top pin is positive, middle is negative, bottom is gauge. I used a multimeter and confirmed positive charge is applied for two seconds at key on.

But at key on, instead of hearing the pump, I see the fuel gauge swing way past full. I'm also getting a SES. What did I miss?

[Rettax3]

Okay, I'm not sure what part of the wiring you've lost. Most replacement pumps do include a small stretch of wiring harness for the inside of the tank, but if yours didn't then it didn't. I would recommend cleaning off the wires from the car with their three-pin connector so that you can see the color-codes and make certain they are matching up correctly. You can also check for fuel-flow: if the positive and negative wires inside the tank to the pump are crossed, and correct voltage is applied to the exterior wires, then the pump will not flow any fuel out of the supply hose at all as it would be running backwards. What did you mean by 'running poorly'? You are right, grey is +12 volts to the pump, black is ground, and purple connects to the gauge sending-unit, which is essentially a variable resistor. If you have to, pull the tank back down to make sure nothing got crossed inside the tank -that is one area where you don't want an electrical short... You can check for continuity across the three pins going into the tank too -the black wire should provide a good clean ground to the body of the tank or sending-unit. Depending on how much fuel is in the tank, there should be some resistance from the black ground wire to the purple sending-unit wire. I don't know what the resistance through the pump should be, but it sounds like it is trying. It sounds like you have the +12 volt to the pump and ground switched to me, but it could just as easily be the other wires crossed too... Be glad you don't have a digital-dash Beretta -they had four wires going into the fuel-tank, not just three!

[Wisconsinite]

Thanks.

I'll double check the pins on the car. I used my AAs to apply power to the new pump. I could hear it running, and when I cranked the engine I could tell it was getting a little gas from pressure in the lines.

I could just crawl under the car and attach my battery pack to the pump every time we drive it. I bet we'd get at least five miles before I had to change the batteries.

By "runs poorly" I mean the old pump made noise with reverse voltage, but didn't pump anything. When wired correctly it gurgled air through the gas left in the lines.

[Wisconsinite]

Progress!

I have the pump wired up correctly now, and the car runs again. I would not say it runs well, though. Better power and acceleration, but it still bogs a bit on the hills and dies at idle.

I'm resetting the PCM to clear the codes. I'll also do an idle relearn. On Monday I'll get the codes read to see which are still an issue. I still can't find a temp sensor to replace the mystery sensor, but I had an idea. I have the wireing diagram, so I know which pin on the PCM it connects to. If I check continuity between that wire and the pins on the suspect sensor I can find out which one it is.

[Wisconsinite]

I've made progress with the engine coolant temp sensor. The wires from the PCM are yellow and purple, and they connect to the top and bottom pins of that 4-pin sensor on the right side of the engine.

I pulled the plug and checked the resistance between those two pins.

With a hot engine (15 minutes after test drive) it was 450 Ohms, which corresponds to 195 degrees F.

After the engine cooled, it was 2200, which corresponds to 86 degrees F. So, it seems like the sensor is providing accurate readings.

Also, why isn't SES light on? Of the 9 codes, most of them should have triggered it. It lights up at key on, so the bulb works.

But the car still runs terribly, so what should I look at next? Can I assume I've resolved the fuel pressure with the new pump, or do I need confirm it with the kit from the store again?

Should I check for vacuum leaks?

It seems to me that the car can idle perfectly fine with throttle applied, but when allowed to idle on its own it can't find a stable RPM and eventually dies.
I tried to relearn the idle, but the car dies immediately in D. Its like the engine doesn't know to increase power when the load increases. What controls that?

[Rettax3]

The Idle Air Control Valve (IAC) controls that, but since you replaced that, it could (in theory) be ruled out. Your throttle-body could be dirty, and the idle-air passages blocked up maybe? We are running out of options to fix on your car... Consider getting a new throttle body gasket, and pulling the TB off for a thorough cleaning.

Since your car doesn't use a MAF (Mass AirFlow Sensor), vacuum leaks would actually cause the idle to be high and unsteady, not typically low and unsteady. If at any point you adjusted the throttle stop-screw, it may now be out of adjustment. Also, although that should not typically need adjustment, if your throttle-body is badly worn, it may be possible that it would need adjustment over time (or a new TB...). If you could ever get a steady idle out of it, you could test the TB by wiggling the throttle-late shaft, and see if the RPMs fluctuate, or if you feel looseness allowing it to wiggle. The main thing is that the computer should be able to set the idle down to the correct minimum idle speed with the engine both cold AND warm. So long as the car doesn't idle too high in neutral/park cold or hot, then you haven't given it too much air with the throttle set-screw. It is exceedingly rare that one of these cars would ever need the set-screw adjusted from original factory settings, unless the engine has been modified or has another problem going on with it, but it could happen. So you are now showing no trouble-codes (SES light) at all?

Edit:
As a related side-note, I had a relative-in-law with a Grand Am replace his fuel-pump last year. After talking him through the procedure for hours over the phone (he is not a mechanic, nor mechanically inclined, but he can actually make vehicles look pretty, which I sort of can't ), he still had problems with the car. He ended up getting another fuel pump, and that fixed it. I don't know if he damaged the first pump, got it contaminated with dirt, hooked it up wrong, or if it was truly defective. But another test shouldn't hurt you.

[Wisconsinite]

I pulled off the throttle body and all attached sensors. Everything inside the TB was coated in a nasty black goo. It was nearly plugging some of the sensor holes, and the butterfly was sticking a bit when closed. I assume that is carbon buildup from exhaust being recirculated by the EGR?

The IAC valve passed Chilton's test. After cleaning the TB I reassembled it and started the car.

I adjusted the idle screw until the car sounded good and didnt shudder and die. It still died when I tried an idle relearn (ran fine for two minutes then died suddenly). But it performed well on a test drive. The acceleration is sluggish (0-60 in 30 seconds) but it didn't die at stops and it did fine on the hills.

This car is much improved, but still has something wrong with it. Perhaps I'll test the ignition modules this evening..

Any other ideas?