Testing a bad engine computer, and its symptoms?

[charlie8575]

I'm about at the end of my rope.

Since the summer, I must've dumped nearly $700 into my car, and it's still not running right, and now that it's getting cold and miserable, and dark early, I don't feel like having it up and die on me in the middle of nowhere.

The car in question is a 1994 Buick Roadmaster, which has an LT-1 in it. There are about 145,000 miles on it.

After changing the distributor cap, it ran like a champ for most of the first part of the year, then it started to get warm.

Since roughly late June/early July, I've done (or had done) the following, all in an attempt to isolate random stalling, usually as the car comes to full operating temperature, and then would still continue to stall sporadically.

-Coolant temperature sensor.

-Thermostat (found to be sticking- also found to be one installed at the factory!).

-Throttle position sensor.

-MAP sensor

-Mass Air Flow sensor.

- Three starters because of all the re-starting I've had to do, which killed two of them.

-A battery (all the stalling did in the other one).

-Coil and ignition control module.

-The fuel pump.

So far, replacing the fuel pump has helped the most. I've been able to drive the car for nearly two hours without trouble, before that, in two hours, I'd probably have had to let it sit for at least 45 minutes and have 15-20 attempts to restart it. Now, if it stalls out, 95% of the time, it re-starts quickly, might die once or twice more, and then continue running.

Prior to the pump replacement, the stalling/restart would also produce variously, or all at once, backfiring, an odor almost like exhaust and coolant mixed with raw gasoline, the smell of straight raw gas, and restart-and-stall after one or two revolutions of the engine.

After two mechanics, a lot of searching to try and isolate what might be going on, and a bunch of other stuff, I've come to the following conclusion, as has my father, who's been a computer engineer for 40+ years:

Dad is thinking the computer might be on the way out.

Evidence:

1. All the problems occur about when the computer should be getting warm. Last night, driving home in cold weather, it took a very long time for problems to develop, and the car was well up to operating temperature. I had the defroster on full, and no heat to the floor. About when the computer probably started getting hot, the problems happened.

2. The difficult re-starting, as Dad described it, is either the computer not reading input from the senors correctly, or the sensors malfunctioning. Well...all the logical sensors have been replaced.

3. The backfire/flooding/strange scents that seem to result from a badly-timed engine or badly-timed electronic components point to the computer not working well.

4. The failure would tend to take a while longer in extreme heat because I'd have the air conditioner on, on days I was operating on the vents alone in warm weather, problems came much sooner.

5. Frequently, the car wouldn't start for 2-3 hours after parking. We're thinking the computer needed time to dissapate the heat.

So....does anyone think we might be on to something- or does anyone have any ideas?

The main question is- is there a way to test the computer to see if it is actually functioning right? I discovered that changing the computer, while simple from a physical R&R standpoint, is a pain because even a rebuilt needs to be programmed and such, something I didn't realize, so it looks like I might have to let a garage do it, but if it doesn't need to be done, I don't want to do it, preferring instead to fix the problem.

Any ideas appreciated.

Charlie Larkin

Edited December 11, 2013 by charlie8575

[slusher]

Charlie taking the battery cable off should clear the computer. Back fireing sounds like timeing or head gasket. Did you change the fuel filter? How is the converter? This is a real good topic for Bill..

[o-man]

Have you been able to extract and codes from computer?

[Eshaver]

Charlie , Usually a G M product such as you're will provide a trouble code . Of course you have to have a diagnostic OBD instrument to get a code out . Now , if a G M car is an OBD II system, 1991 and newer, it even gives a lot more specific information that my ANTIQUE 1987 Oldsmobile will not . I found I had a Crank Position sensor bad and a bad coil pack even though the computers used then will not provide such information . I'm sharing this because many of the symptoms you are experiencing, I had too. Oh, the Oldsmobile I have is seldom driven . I only have 78,000 from the time it was new .

Now if you really want to just do a "Back Yard " test , as it's been mentioned , first dis connect the Ground on the battery and wait five minutes . If you're still experiencing the same symptoms , I have actually gone to a Pull you're own part yard and bought a "Like " computer from a similarly equipped car . Yes, a used computer will run a car that is not necessarily equipped just the same way . Chrysler on the other hand ............... Hummmmmmmm, don't get me started on anything Chrysler anymore !

Edited December 11, 2013 by Eshaver

[charlie8575]

Charlie taking the battery cable off should clear the computer. Back fireing sounds like timeing or head gasket. Did you change the fuel filter? How is the converter? This is a real good topic for Bill..

Timing had crossed my mind, but where it runs very smoothly and without hesitation when it's running (and still has plenty of power for high-speed merges), I'm leaning towards "no," but it stays in the back of my mind.

The cats, from what I can tell, are good. I haven't done a vacuum pull, but the exhaust seems to flow freely, so I would think they're not clogged.

Fuel filter was changed with the pump.

I've seen no evidence of blown head gaskets on the radiator or oil cap. Doesn't mean it didn't happen, but I haven't see it. Since changing the thermostat, the car is running considerably cooler, so I'm tending to think they're okay.

Charlie Larkin

Have you been able to extract and codes from computer?

Not yet, Oscar. Occasionally after stalling, the Service Engine Soon light will come on for a minute or two, but it doesn't seem to be storing anything.

Charlie Larkin

[charlie8575]

Charlie , Usually a G M product such as you're will provide a trouble code . Of course you have to have a diagnostic OBD instrument to get a code out . Now , if a G M car is an OBD II system, 1991 and newer, it even gives a lot more specific information that my ANTIQUE 1987 Oldsmobile will not . I found I had a Crank Position sensor bad and a bad coil pack even though the computers used then will not provide such information . I'm sharing this because many of the symptoms you are experiencing, I had too. Oh, the Oldsmobile I have is seldom driven . I only have 78,000 from the time it was new .

Now if you really want to just do a "Back Yard " test , as it's been mentioned , first dis connect the Ground on the battery and wait five minutes . If you're still experiencing the same symptoms , I have actually gone to a Pull you're own part yard and bought a "Like " computer from a similarly equipped car . Yes, a used computer will run a car that is not necessarily equipped just the same way . Chrysler on the other hand ............... Hummmmmmmm, don't get me started on anything Chrysler anymore !

Actually, Ed, this one is still OBD-I, as I can decode it with a paperclip! (Well, they did one thing right...)

My car doesn't appear to have a crank sensor- I think that was mostly used with the distributor-less ignitions on the EFI V/6s and some of the Cadillac V/8s.

Coil....you did remind me. I did change the coil and the ignition control module, too. Thanks for the memory jog! I'll edit that now above to make sure it doesn't get missed.

Charlie Larkin

[o-man]

Charlie, does it have climate control? If so, you can access codes through it.

[slusher]

Charlie it could be a small amount of antifreeze going in to a cylinder and would slowly lose coolant. The smell is coming from some where and no heat sounds like a coolant issue...

[1972coronet]

I'm thinking it's the Crank Position Sensor as well .

OBD II didn't appear until the 1996 model year .

[gluebomb]

Those Optisparks sucked when they got hot, wet, old, or if you looked at it funny, they act up. An Optispark would be the whole distributor. OBD-2 started showing up on some 1995 model year cars. It really sounds like you may want to bring to a Chevy Dealership. If you think it is the computer I would trace all wires ohm them out, check voltages, make sure nothing is broken.

Edited December 11, 2013 by gluebomb

[kennb]

Check the fuel line from the tank to the engine,,,,,,,,,,,there should be an additional filter under the passenger door bolted to the frame. It sounds more like a clog in the fuel line than anything else.........I am assuming you have replaced the plugs also............

[lizardlust]

Also, if my memory serves me right the exhaust gas recirculation(EGR) valves in the LT1 motors were prone to sticking open after 50k miles, causing symptoms similar to a massive vacuum leak. Has this been replaced? I know of more than one GM B-body that was sent to a recycler due to this........

[Guest]

Take it to a dealership and have a diagnstics done on it. That way, you won't have to guess what's wrong with it, you'll know exactly what's wrong with it. You surely aren't going to find the answer to your car's problem on a model car forum. Anyone can guess what the problem is. Look at all of the money you've spent already trying to figure out what's wrong.

Several years ago, my mother was driving a '94 Cougar. It developed a problem where the transmission would act like it went into neutral. You could let off of the gas pedal and you could feel it go back into gear. It didn't didn't do it often at first. But the frenquency began to increase. The car had less than 50,000 miles on it. So, I was pretty sure the transmission wasn't bad. I told my father to take it to a dealership and for $42, they would hook it up to a computer and find out what's wrong with it. He wouldn't do it. Said he wasn't going to pay $42 for them to just tell him what was wrong with it. He took it to a transmission shop. Their diagnosis was the transmission needs to be rebuilt. Imagine that! Twelve hundred dollars and three months later (exactly when the warranty went out) it started doing it again. I told him again, "take it to a dealership and find out what's wrong with it!". He finally gave in and took it to a local Ford dealership to get a diagnostics done on it. The problem? It was a neutral safety switch that was still somehow an original part. Hmmmm... The cost? $120 dollars total. That was for the diagnostics, part and labor. The transmission hasn't missed a beat since and it has around 160,000 on it.

[Ace-Garageguy]

Also, if my memory serves me right the exhaust gas recirculation(EGR) valves in the LT1 motors were prone to sticking open after 50k miles, causing symptoms similar to a massive vacuum leak. Has this been replaced? I know of more than one GM B-body that was sent to a recycler due to this........

1) This is a valid point. The '92 Chevy CK1500 truck (iron smallblock Chevy 350) I currently own had some odd problems before I bought it. The then-owner (a good friend) had taken it to several "diagnostic specialists" like Jiffy Tune, and every one told him it needed a computer. Because it WAS running occasionally like it had a massive vacuum leak, and would sometimes stall and refuse to hot-start (and because I hadn't gotten into electronic diagnostics myself yet), I convinced him to replace the EGR valve...$14 at the time, I think...before he sprung for a computer. Bingo. No more problem. Still running fine now with over 260, 000 on the clock.

2) How did you come to replace all of those parts? Throwing parts at a vehicle rarely accomplishes anything other than wasting money. Did you or someone else pull codes from the OBD I before replacing all of that stuff? OBD I was on vehicles prior to 1996, though some earlier vehicles have OBD II style interface connectors., and GM has some OBD "1.5" systems back to 1994. Geez.

3) There are other diagnostic procedures for that vehicle (and all OBD I systems) that do not require sophisticated computer interface capability. They can be found in the factory service manual, and usually require only a good volt /ohm meter with the appropriate probes, and possibly a fuel-pressure gauge (with appropriate adapters to hook up to high-pressure EFI). The steps must be followed carefully, in order, in order to generate any valid information (which is why most mechanics won't do them). Factory books are available on the 'Bay. Last one I bought for an '89 Celica GT, with complex electronic issues, was only $50, used. Isolated and corrected all the problems, including rodent-damaged wiring and bad relays from sitting for years, in just a few hours.

4) Trying to diagnose a problem, particularly an INTERMITTENT problem over the internet is most likely an exercise in futility. An intermittent problem COULD indeed be a failing computer. It could also be a corroded connector in the wiring harness somewhere, a bad connector on any of the sensors, an intermittent bad ground, a failing intake-air-temp sensor, or even foreign matter in the fuel tank.

I encountered a dual-throttle-body EFI Corvette years ago with similar problems to yours. It would run fine cold, but as it warmed up after being driven, would begin to buck and spit, finally quitting entirely. After a lengthy "cool down" period, it would run fine again. Once more, all the usual suspect parts were thrown at the car to no avail. When it came to me finally, I went through all the manual diagnostic procedures, including hooking up a fuel-pressure gauge I could read while driving. Surprise !!! Fuel pressure would drop off after a while, finally to zero. Car would stall. Nursed it back to the shop and checked the filter and lines...fine. I don't recall the fuel pump location, but it tested fine. Hmmmm....

What I finally found was sugar in the tank. Sugar does NOT dissolve in gas, but as the car is driven, it would progressively coat the screen-sock over the fuel pickup, and eventually stop the flow. Once the engine was shut down (to "cool off"), the sugar would gradually drop off of the screen and the engine would once again run fine.

THE POINT: Model car forum diagnostics probably won't solve your problem. There are simply too many variables. I suggest either you get the factory book and go through the procedures yourself, or find a VERY competent mechanic who will. The dealer can't pull any more info from the OBD system than you can (other than things like FI pulse width and duty cycle) and...surprise...not all dealer mechanics are competent.

If it turns out you DO need a computer, I just bought a rebuilt one for a '96 PT Cruiser from these folks. The original unit had an intermittent "internal-check-sum-failure". The company that supplied the new one (OBD II in this case) programmed the mileage and VIN before shipping. Ran perfect right out of the box...all for less that $170. http://stores.ebay.com/autocomputerexchange/

Good luck.

PS. Don't bother suspecting things like blown head gaskets, etc. without doing compression tests on every cylinder, and preferably leakdown tests as well, and a combustion-gas-in-coolant test. Intermittent drivabliity issues are not caused by severe internal engine issues...usually.

AND...if replacing the fuel pump made the most improvement, see if the vehicle has a separate fuel FILTER somewhere. Your problem COULD be similar to what happened in the Corvette above BUT...in your case, it MIGHT be from scoring inside the old pump, from pushing rust particles through it, progressively weakened the pump, and as you drove, the filter would become progressively blocked to the point of engine shut down. Replacing the pump, as you did, with one that can make full pressure may be enough to overcome a partially blocked filter...until it becomes fully blocked again. Just wild speculation based on symptoms.

PPS. I THINK your engine computer gets its crankshaft position signal from a module in the distributor body.

Edited December 11, 2013 by Ace-Garageguy

[DrKerry]

Computer could be part of it. I had a 92 Mits Galant that when colder air or foggy weather the car would run rough and quit. It'd take ten minutes of cranking to get it to re start only to get up to 20 mph to have it stall again. It's made my hour and twenty minute ride up to three hours to get home. Trust me I was glad to see it go when it did. Anyway, what I had found out the first time when I was trucking down the highway at 70 it dies at speed. So trying to decifer the problem and going through replaced timing belt and pulled the head to inspect the valves. I put it back together and it never ran right again. I replaced the compuer with a 500.00 rebuilt one and it still ran like BLAH_BLAH_BLAH_BLAH. I'd get going and it would quit. so I started to go through everything. Ignition module, distributer, coil, plugs, wires, throttle position sensor, idle speed sensor, throttle body, mass air flow sensor and nothing. Finally I was done with it, it got parked and sat all winter. Finally in the spring one last hoorah I posted for help on Craigslist looking for a used computer just so I could troubleshoot it and see and a guy an hour North of me came back saying he had one. He drove down, we plugged it in and after 3000.00 dollars worth of work and labor it came down to an ill rebuilt computer which by then was to late to send back. So I sent it to someone my friend knew and 25.00 later the car was running like a top.

Only to have the frame rust out the following year...... Oh well.

Check the computer on it and see if that may be the problem... Good luck!!!

[Ace-Garageguy]

Hold the car up to the computer and let me look at it. Stupid, right? So is speculation.

The only computer-specific codes I've found for pre-OBD II GM cars are 47,51, 55, 85, 86, 87. Any of these COULD indicate internal failure.

BUT IT DOESN'T ALWAYS WORK RIGHT, especially on older vehicles, and if it's intermittent, you need to try to pull the codes while it's malfunctioning. Are you getting a "check engine" light while any of this is happening, by the way?

I have an 89 EFI, OBD I, GMC 305 cu.in. longbed shop-truck as well. Runs like ###### sometimes, runs fine other times. NO CODES. It's just old and confused. Intermittent problems require actual hands-on diagnostic work, which I haven't bothered to do yet.

Edited December 11, 2013 by Ace-Garageguy

[Ace-Garageguy]

Computer could be part of it. I had a 92 Mits Galant that when colder air or foggy weather the car would run rough and quit. It'd take ten minutes of cranking to get it to re start only to get up to 20 mph to have it stall again. It's made my hour and twenty minute ride up to three hours to get home. Trust me I was glad to see it go when it did. Anyway, what I had found out the first time when I was trucking down the highway at 70 it dies at speed. So trying to decifer the problem and going through replaced timing belt and pulled the head to inspect the valves. I put it back together and it never ran right again. I replaced the compuer with a 500.00 rebuilt one and it still ran like ######. I'd get going and it would quit. so I started to go through everything. Ignition module, distributer, coil, plugs, wires, throttle position sensor, idle speed sensor, throttle body, mass air flow sensor and nothing. Finally I was done with it, it got parked and sat all winter. Finally in the spring one last hoorah I posted for help on Craigslist looking for a used computer just so I could troubleshoot it and see and a guy an hour North of me came back saying he had one. He drove down, we plugged it in and after 3000.00 dollars worth of work and labor it came down to an ill rebuilt computer which by then was to late to send back. So I sent it to someone my friend knew and 25.00 later the car was running like a top.

Only to have the frame rust out the following year...... Oh well.

Check the computer on it and see if that may be the problem... Good luck!!!

Not meaning to be an SOB, but all of that "3000 dollars worth of work" could have been avoided if competent diagnostics had been done FIRST.

I see this CONSTANTLY.

Let's say you were always out of breath and had frequent headaches. What would you think of a doctor who walked around you, thumped a couple of things, and said "you need a lung transplant"? Same thing.

Edited December 11, 2013 by Ace-Garageguy

[lordairgtar]

NO CODES. It's just old and confused.

Thats it!!!

[my80malibu]

In 1994 GM had most of the computer bugs worked out, that had caused them issues in the mid to late eighties. The PCM used in a 94 is a lot more robust computer than before. Charlie do you have the Federal Ignition Package?. Most of the Caprices, that I worked on back in 94 had the distributor cap behind the water pump, these were Police cars. 5.0L with LT 1 ignition system. Does your car have that type, or does it have a actual distributor at the rear? The proble you have been describing sounds eerily similar to one that I had, it turned out to be a Pickup coil in the distributor.

Very easy to test, those wires become corroded at the base where they go into the ciruit, they dont always brake but cause a lot of headache. You would need to OHM out the wires and compare it with the spec, in a repair manual like the ones you get at the parts store.

The Early computers you could take out of the mounting fixture, hold it in your hands, and while the car was running you could gently squeeze the case, it would sometimes change the way a vehicle would run like having a bad connection. I found one this way in a buick century it also had intermittant issues, like No AC, no Hard code was found, but several soft codes were stored.

Edited December 12, 2013 by my80malibu

[von Zipper]

Being a 20 year old car from the snow/rust belt it wouldn't hurt to take the negative cable apart at the engine block and wire brush the connections, from the engine block to the frame also. It may not fix the problem but can't hurt any thing either-I've seen bad ground wires do weird things.

[farmer1]

Just because you replaced a part, like the fuel pump for example, doesn't mean you put a good part back on. The quality of much of the parts store replacement parts is poor and getting worse. I am so tired of replacing parts over and over and sometimes losing customers. Lately I am seeing their parts warranties getting shorter and cutting or eliminating labor claims too, kinda proves my point doesn't it.

Good luck with it Charlie.

[DrKerry]

Not meaning to be an SOB, but all of that "3000 dollars worth of work" could have been avoided if competent diagnostics had been done FIRST.

I see this CONSTANTLY.

Let's say you were always out of breath and had frequent headaches. What would you think of a doctor who walked around you, thumped a couple of things, and said "you need a lung transplant"? Same thing.

Likewise not trying to be an S.O.B. but it went to a Mits dealer twice as well as a local shop in my area. No one knew what was happening, no codes thrown. Nothing but the headaches of how it was running. I've been a mechanic for 30 plus years, I've been under the hood a few times and when it stumped me I went to the dealer, when it stumped them I gave up. They told me one time that it was transmission realated to change the fluid and filter and put in a Mits additive to the trans fluid...... Ahh yeah!!!! Trust me it wouldn't of gotten over a thousand dollars if I didn't have a persistant wife saying it can be fixed, it can be fixed.. I would of scraped it along time before all of the mess it created.

In the end it was a blown out capacitor in the computer that was leaking fluid, in the end it leaked out ran down the pc board and ran itno the capacitor that runs the fuel pump. Tests ran at dealership didn't bring the info up that would indicate a computer problem at that time.

[Ace-Garageguy]

Just because you replaced a part, like the fuel pump for example, doesn't mean you put a good part back on. The quality of much of the parts store replacement parts is poor and getting worse. I am so tired of replacing parts over and over and sometimes losing customers. Lately I am seeing their parts warranties getting shorter and cutting or eliminating labor claims too, kinda proves my point doesn't it.

Good luck with it Charlie.

Yup. I recently got 3 bad Mexican-rebuilt master cylinders for an old ('86) Ford truck in a row. They just wouldn't make pressure. Finally in anger and frustration, I took the last one apart to see what the problem was. The seals were in BACKWARDS. Impressive craftsmanship and quality control. I found a NOS OEM seal kit online and rebuilt the old one myself. Perfect.

I've also had a rash of supposedly quality aftermarket replacement parts for fairly recent Mopars lately. Chinese radiators with the mounting holes in the wrong places, engine mounts that had to be modified to even fit the vehicle, and most recently, "exact replacement" suspension lower control arms that were CAST aluminum instead of FORGED. Dangerous, and I refused to install them.

The very high percentage I get of BLAH_BLAH_BLAH_BLAH parts and vehicles referred to me after the owners have been told "there's nothing wrong with it...it just won't run right" (by professional "mechanics") tells me that there's something very wrong in the car repair industry.

[DrKerry]

Yup. I recently got 3 bad Mexican-rebuilt master cylinders for an old ('86) Ford truck in a row. They just wouldn't make pressure. Finally in anger and frustration, I took the last one apart to see what the problem was. The seals were in BACKWARDS. Impressive craftsmanship and quality control. I found a NOS OEM seal kit online and rebuilt the old one myself. Perfect.

I've also had a rash of supposedly quality aftermarket replacement parts for fairly recent Mopars lately. Chinese radiators with the mounting holes in the wrong places, engine mounts that had to be modified to even fit the vehicle, and most recently, "exact replacement" suspension lower control arms that were CAST aluminum instead of FORGED. Dangerous, and I refused to install them.

The very high percentage I get of ###### parts and vehicles referred to me after the owners have been told "there's nothing wrong with it...it just won't run right" (by professional "mechanics") tells me that there's something very wrong in the car repair industry.

I worked at Autozone for about two years, you wouldn't beleive how many altenators we got back because they were put together wrong....

[Ace-Garageguy]

I worked at Autozone for about two years, you wouldn't beleive how many altenators we got back because they were put together wrong....

Yes sir, I believe it.

Kinda makes you question the wisdom of sending all these parts out of the country to get rebuilt. I'd MUCH rather pay an extra $20 to get an alternator that was rebuilt in the USA and worked right the FIRST time, instead of getting garbage, installing it, only to find it's BLAH_BLAH_BLAH_BLAH and having to go back to try another one.

Again, there's nobody driving the bus who has a clue.