What's wrong with my 2.8?

[89Hatchback]

My '89 Corsica is giving me fits. "The problem" announced itself one day - When I went to start it, it cranked, caught like normal, but only ran for a second before cleanly dying. (as in the idle fell rapidly until it died - no struggle) This repeated several times. I finally turned it completely off. I can't remember if it stayed running after that or not, but I finally did get it running, and it was alright.

Then it manifested itself as a few troubling burps while driving.

One night I was able to cox it home with some extra throttle, but most of the times when it's having an "episode" while starting, the throttle has zero effect.

One time I could feel it acting up while coasting. Sure enough, when I got it home it idled like crap. It got worse. Slowed way down (with an a hissing noise getting louder as it died (huh?), then caught itself and raced, found a good RPM but ran with a distinct miss. It did finally throw a code during the poor idle: 33 and 44.


Most of the times the episodes seemed short lived and it starts running fine after a few minutes. But last Friday it gave me fits at two of my stops and finally died heading home. I couldn't keep it running for more then a few seconds and had it towed to a mechanic.

Of course the mechanic didn't work in it until Monday. The guy's had it for 4 days now and it's mostly behaving itself - not good for diagnosis.

The only thing I know for sure is I've got good fuel pressure on the rail when it's acting up. I'm leaning toward a flaky computer. (Another shop changed it back in November on a bad diagnosis. Is $135 parts cost suspiciously cheap?) Part of the reason is that it seems like it improves if I power off first instead of starting it from "run". But I'm also wondering if I've got a bad injector shorting the whole thing out (the engine is wired for bank-fire, but it all goes to one terminal on the computer.)

Of course I can't eliminate spark entirely, but I haven't smelled any fuel.

The tach indications seem normal (considering) so I don't think it's the crank sensor (went though that a year or so ago.)

If I'm giving mixed symptoms, it's because I can't really find a pattern. Just when I think I've found one, it goes and proves me wrong.

Ideas, please. Before I throw parts (or expendable heavy objects) at it.

[ssdurango]

EGR valve plugged would be my best guess.

[89Hatchback]

EGR valve plugged would be my best guess.

Ok, that wasn't on my list. I'm not sure how it would be causing the problem. I didn't think the EGR opened up until "underway" and wasn't a factor at idle (unless stuck open). Am I wrong?

I should add that the car will die under power (when it wants to).

One thing is consistent - it always catches when I try to start. It just won't always stay running

[89Hatchback]

Sorry, I haven't dealt with stuck EGR before - are you saying the device that controls the diagram for the EGR is stuck such that the EGR opens anytime there's engine vacuum? I guess that could explain the series of start/die episodes. Seems surprising that it could kill the engine while under power.

[89Hatchback]

Well, I tried removing the vacuum hose from the EGR so it can't open. No luck. It still stalls when it wants to.

[86FieroSEv6]

Your symptoms sound like crank sensor or MAP sensor issues to me. You could pull down the ECM and have somebody gently rap on it and flex the case while you are driving as well.

[89Hatchback]

Your symptoms sound like crank sensor or MAP sensor issues to me.

I replaced the crank a couple of years ago. I'd hope it's not going out already. When mine died, it gave a few warnings, and then went solid.

Today marks two weeks since the onset of symptoms.

Not sure how a MAP fails - that would be a new experience for me. The mechanic I talked to said they tend to go bad and stay bad. How can I test it?

I'll be trying the "play with the wiring" today.

[86FieroSEv6]

Yeah, that's what I thought too. My baby 'burban (94 S10 Blazer) was giving me the same symptoms. I replaced my ign. mod., fuel pump, injector assy., ign. coil and pulled out several hairs. Every one of those things made my truck run better, but the root symptom remained. Then, on a whim, I grabbed a MAP off one of the parts cars and plugged it in . . . . . . . . . . . . and I'll be damned if that wasn't the main problem all along. Really weird, 'cause there was no obvious signal fault on the scanner. Are you getting any trouble codes?

[89Hatchback]

Are you getting any trouble codes?

One time when it was idling badly I saw the light come on so I checked codes: got a 33 and a 44. The book indicated that a 33 could be thrown by idling problems. 44 is oxygen sensor showing "lean".

My wiggle test yields nothing.

When it gave me fits on starting, I did unplug the MAP in hopes of limping home. No change, nor improvement. I thought the car was supposed to be able to limp, but I'm not sure what it needs to detect that the MAP is inop.

I should add is that I've never changed the MAP, so it's at least 12 years old - maybe even original. I suppose that might be asking a lot for a mechanical sensor.

[86FieroSEv6]

yep

Uh-huh . . . that's exactly what I did, and with the same results. My studies caused me to assume that even in backup spark and fuel control, the MAP is apparently one of the critical inputs for base operation along with the coolant temp sensor. So, my understanding is that for our particular ECM's there some things that the system cannot assume. Of course, with OBDII, all parameters can be simulated for limp in strategy. Can you find somebody with an analog (live) type oscilloscope and perform a sweep test of your throttle position sensor, or have you tried to run with it disconnected.

[89Hatchback]

I really liked the MAP sensor idea. But it still died on me under power after I replaced it.

I checked the TPS. I couldn't find the probe to my scope, but I used my meter or both DC and AC and didn't find any noise or "rough spot".

The mystery continues.

[86FieroSEv6]

It might be time to take a closer look at the fuel pump. Have you monitored the current draw of the fuel pump? Have you checked the fuel pump relay socket terminals for heat scorching or looseness? Also, in our year models GM had fuel pump connector issues and many replacement fuel pump modules come with updated connector/pigtail kits.

[89Hatchback]

My mechanic thinks he found a problem - the ignition switch. It got to the point it wouldn't start (or should I say - stay running once caught) so I had to have it towed in. He said that if you didn't let the key back all the way into run, it would continue running after starting. However, I still have a stalling issue. So either "that wasn't it" or it's one problem down, one to go. So, in some ways, I'm back to square one - it stalls (or tries to) when driving.

It might be time to take a closer look at the fuel pump. Have you monitored the current draw of the fuel pump? Have you checked the fuel pump relay socket terminals for heat scorching or looseness? Also, in our year models GM had fuel pump connector issues and many replacement fuel pump modules come with updated connector/pigtail kits.

Thanks for the ideas. I sure wish the line on my fuel pressure gauge was longer. There's no way I can see it with the hood closed. So far, since changing the ignition switch, I've not had a starting problem. Both me and my mechanic have found the pressure to be good. I did verify it was good when I was having starting problems. But the stalling might be a completely different thing. But the only way I can run tests is to get an indicator where I can see it while I'm driving.

The fuel pump relay shouldn't be an issue except for starting or if I've lost the oil pressure switch. Checking the connector by the pump is a good idea. It seems most likely to die on a right turn, but it can do that other times as well. So far everything has passed the "wiggle test", but I didn't try wiggling anything back by the pump.

I don't remember seeing any pigtails when I replaced the pump some years ago.

[89Hatchback]

Well, now I really am back at square one. Went to get into the car for the drive home, and it was back to the crank-catch-die game. Since it was cooperative, I did some checks. Once again, good fuel pressure. Using a test light, there's power to the ECM. The injectors fire, but they seem to stop just before the engine dies. Oddly it seems to give a "last shot" just as it the engine finally stops. Probing the "distributor" inputs with my volt meter, it appears that the signals from the ignition system is there until the bitter end (and then some - perhaps slow meters). So at this point it's looking like a ECM problem. I did try swapping the MAP sensor - no change. I haven't pulled the plenum to verify the vacuum hoses under it, but I'm having a hard time figuring out how any leak could cause the problems I've been having - especially a sudden die upon acceleration. Everything I can touch seems to pass the wiggle test. The TPS was previously tested and seemed OK. I do see the injectors not fire when I try to start in clear flood mode. But otherwise there's no apparent response to changes in the throttle position. So I'm not coming up with anything but the computer. (I sure hope it's not the memcal.....).

This ECM was (erroneously) changed in November. I'm wondering if I inherited someone else's problem child.

[86FieroSEv6]

Was the symptom present before the ECM swap?