TBI woes

[66gmcguy]

Hello folks. I normally post over in the 73-87 section, but since I have a TBI setup in my 1982 GMC, I figured I'd get more help over here, since 1987 was the first year for TBI trucks.

A few months ago I bought a 1987 Chevy R20 with a TBI 350 and manual transmission to use as a parts truck for my 1982 GMC. It seemed to run okay but the throttle was extremely sticky due to the flappers in the throttle body sticking. I tore down the TBI, bushed the throttle shafts, replaced the injectors and IAC valve (both with Standard Motor Products), and did a thorough cleaning, making sure to clean all vacuum ports and the IAC passage. I replaced every gasket, seal, and the FPR diaphragm. After doing so, I've never really been able to get this thing to run right since then. Here are a few problems I can't get ironed out. I have WinALDL and no codes are present.

At idle, the RPM fluctuates up and down. If I depress the throttle a little bit, the RPMs still fluctuate, just in a higher RPM range than in idle. I know this is either due to low fuel pressure, a vacuum leak, an IAC problem, or TPS problem. Thing is, I have a new fuel pump + hose, all new fuel lines, fuel filter, all new intake gaskets, and new vacuum lines. I've checked fuel pressure (11.5 PSI), and checked for vacuum leaks by spraying starter fluid around every possible source for a leak and found nothing. I've tried plugging every vacuum port with no luck. I have the timing set at 0 degrees (with the bypass unplugged). I must have done the IAC reset procedure 50 times, I've put the old injectors and IAC back on, replaced the TPS and set it to .5 volts at idle, all with no change. The thing is, if I turn the idle screw up so it revs around 900 at idle, it seems to stabilize (only when in closed loop), but the IAC counts are zero. Another thing I can do that seems to help is unplug the CTS which richens the A/F mixture. I can't tell if it's the richer mixture or the slight raise in RPM that's fixing the problem though. When the engine is cold it idles fine until the coolant temp reaches maybe 100ish degrees, then I get the fluctuating idle (until it goes into closed loop mode.)

Another thing is when accelerating, I have a very distinct, but momentary misfire at a certain RPM. It only happens under light to moderate acceleration though, right at around 2200-2300 RPM, and it's only for a split second. The best way to describe it is it feels like the engine shuts off for a split second then turns back on almost instantly. It doesn't feel like a typical lean-sputter, or backfire. I thought maybe this was a TPS problem, but it does it on all three TPS sensors that I've tried, all at the same RPM under same conditions.

The last thing. When the engine is cold, it drives like crap and has no power it feels like, and every now and then I get a CEL but with no codes. If I shut the engine off and shut it back on right away, the CEL turns off, and it seems to be fine after that. None of this happens after it's warmed up beyond maybe 120-150 degrees. As I mentioned above, it idles fine when cold.

Since having these problems, I've replaced the TPS, MAP, CTS, knock sensor, EGR valve, PCV valve, O2 sensor (3 wire unit), wire harness (Painful Performance Swap harness), distributor (and all internal components), spark plugs, wires, every single intake/exhaust gasket, etc. Pretty much everything short of the EGR solenoid, the ESC, and the PROM. All components used apart from the IAC and injectors are all ACdelco. I've even switched out the computer from another truck that I have with no luck.

I know this is long winded, but I am genuinely stumped, and would love some input. As mentioned, I have WinALDL and can datalog and post it on here if needed. I'm also open to the possibility that I did something wrong by either not doing a test/reset procedure properly or didn't install something correctly, etc, so if you think I should recheck something that I've already messed with, just let me know.

[special-K]

I wish I could help but you already know more about the TBI than I do. I've never had to service that on any vehicle I've had with it. I'm sure some insight will be coming along

[Second Series]

Originally Posted by special-K

I wish I could help but you already know more about the TBI than I do.

Same here...
How about the temp sender? My temp sender worked, but the range was off. Also did you try to induce a vacuum leak by pulling a line to see how it affected things? This sounds like a TBI swap, how is the wiring, maybe a loose ground? Just grasping at straws here...

[Just call me Sean]

If the light comes on there should be a code. Maybe the computer isn't getting constant power for the memory.

[66gmcguy]

Originally Posted by Just call me Sean

If the light comes on there should be a code. Maybe the computer isn't getting constant power for the memory.

I'm not sure if that's the case. I only think this because the computer retains fault codes even after shutting the truck off for a while and the codes only go away after disconnecting the battery for a minute or so.

[66gmcguy]

Originally Posted by Second Series

Same here...
How about the temp sender? My temp sender worked, but the range was off. Also did you try to induce a vacuum leak by pulling a line to see how it affected things? This sounds like a TBI swap, how is the wiring, maybe a loose ground? Just grasping at straws here...

All grounds are good. All ground connections are bare metal and tested with a multi meter.

As far as the rest of the wiring is concerned, I don't believe that to be the issue, since all of the same problems occurred when the engine was still in the donor vehicle with the old harness. I have the same problems even with my new harness.

I will check the CTS tonight and report back.

[ChevyTech]

Make sure the PROM (chip) in the ECM matches your engine size.

[66gmcguy]

Originally Posted by ChevyTech

Make sure the PROM (chip) in the ECM matches your engine size.

It does. The code is ACTC

[Just call me Sean]

Originally Posted by 66gmcguy

I'm not sure if that's the case. I only think this because the computer retains fault codes even after shutting the truck off for a while and the codes only go away after disconnecting the battery for a minute or so.

The computer should retain the codes but it won't if you don't have constant power to it. So it gets power to run but shutting it off removes the only power it gets, pretty much like disconnecting the battery. You said the light comes on but no codes stored.

[Warrens69GMC]

Post here, they will help you figure it out

https://www.thirdgen.org/forums/tbi/

[66gmcguy]

Originally Posted by Just call me Sean

The computer should retain the codes but it won't if you don't have constant power to it. So it gets power to run but shutting it off removes the only power it gets, pretty much like disconnecting the battery. You said the light comes on but no codes stored.

My apologies, I was wrong in my original post. The CEL came on because of a lean code (code 44.) Long story why I originally thought that no codes were present when the CEL came on.

I just drove it around again after disconnecting the battery to wipe the memory and clear any codes, and sure enough, the CEL came on again for code 44. If I shut the truck off and shut it back on right away, the CEL turns off. Like I say, this situation only happens if I drive the truck before it fully warms up. If I let it warm up before driving it, this doesn't happen.

So to circle back, there IS constant power to the computer even with the key off, as it does store codes. so that doesn't seem to be the problem.

I should emphasize that I don't get that code at all if I let the truck warm up before driving it, AND after I've driven it around under various types of loads (I have city/mountain/highway driving everyday during my commute.)

So the lean condition probably explains the poor drivability with a cold engine, but I don't think it explains the other two problems I'm having, because even when the idle fluctuates around, I don't get any codes.

[ChevyTech]

It set a lean code more then once and a lean condition fits some symptoms.

Is the three wire oxygen sensor a Bosch unit? If so I suspect it is the problem.

I would try unplugging the oxygen sensor and seeing how it runs as it warms up.

If the system is seeing oxygen sensor readings that show the sensor is warm enough to use, it will be reaching closed loop operation when the engine is not as warm as it would be if it had a one wire oxygen sensor.

Let us know how it runs with the oxygen sensor unplugged.

[66gmcguy]

Originally Posted by ChevyTech

It set a lean code more then once and a lean condition fits some symptoms.

Is the three wire oxygen sensor a Bosch unit? If so I suspect it is the problem.

I would try unplugging the oxygen sensor and seeing how it runs as it warms up.

If the system is seeing oxygen sensor readings that show the sensor is warm enough to use, it will be reaching closed loop operation when the engine is not as warm as it would be if it had a one wire oxygen sensor.

Let us know how it runs with the oxygen sensor unplugged.

It is an ACdelco unit. Part number AFS74.

Interesting theory. I actually thought about that last night as well. I was going to try it this morning before driving to work but I didn't get a chance to. I will try it this afternoon and report back.

Do you have any thoughts as to what might be causing this problem? (See below)

Originally Posted by 66gmcguy

Another thing is when accelerating, I have a very distinct, but momentary misfire at a certain RPM. It only happens under light to moderate acceleration though, right at around 2200-2300 RPM, and it's only for a split second. The best way to describe it is it feels like the engine shuts off for a split second then turns back on almost instantly. It doesn't feel like a typical lean-sputter, or backfire. I thought maybe this was a TPS problem, but it does it on all three TPS sensors that I've tried, all at the same RPM under same conditions.

This one bugs me the most for some reason. Could it possibly be a defective PROM? That's pretty much the only variable that I haven't eliminated at this point. I have a datalog that shows it happening, but me trying to understand datalogs is like a monkey trying to make sense of an iPad.

[ChevyTech]

Originally Posted by 66gmcguy

Another thing is when accelerating, I have a very distinct, but momentary misfire at a certain RPM. It only happens under light to moderate acceleration though, right at around 2200-2300 RPM, and it's only for a split second. The best way to describe it is it feels like the engine shuts off for a split second then turns back on almost instantly. It doesn't feel like a typical lean-sputter, or backfire. I thought maybe this was a TPS problem, but it does it on all three TPS sensors that I've tried, all at the same RPM under same conditions.

You have not mentioned anything about the distributor yet.
I would try a different Ignition Control Module in the distributor.
It sounds to me like the module could be switching back to the base timing mode for a second.
When you are in the distributor make sure the distributor shaft is not loose and can wiggle sideways.

These distributors also have problems with the shaft losing its magnetism but that usually shows up as a hard start or problems at lower RPMs.

A Cracked insulator on a spark plug can cause that same sensation. Also a torn plug wire boot. I usually could feel this best at a very light acceleration at about 40 MPH

[66gmcguy]

Originally Posted by ChevyTech

You have not mentioned anything about the distributor yet.
I would try a different Ignition Control Module in the distributor.
It sounds to me like the module could be switching back to the base timing mode for a second.
When you are in the distributor make sure the distributor shaft is not loose and can wiggle sideways.

These distributors also have problems with the shaft losing its magnetism but that usually shows up as a hard start or problems at lower RPMs.

A Cracked insulator on a spark plug can cause that same sensation. Also a torn plug wire boot. I usually could feel this best at a very light acceleration at about 40 MPH

It's a brand new distributor, and the shaft is nice and tight. It is a Cardone unit, but after experiencing the same symptoms as the old distributor, I replaced the ICM and pickup coil with ACdelco parts. Still no change from before. Spark plug wires are new, again with no change from before.

Could the ESC relay (or whatever the part is called that the knock sensor connects to) be causing this? I'm sort of doubtful of this since I swapped one over from another truck and it still did it, but I guess I shouldn't assume that that one was okay either.

[ChevyTech]

Originally Posted by 66gmcguy

It's a brand new distributor, and the shaft is nice and tight. It is a Cardone unit, but after experiencing the same symptoms as the old distributor, I replaced the ICM and pickup coil with ACdelco parts. Still no change from before. Spark plug wires are new, again with no change from before.

Could the ESC relay (or whatever the part is called that the knock sensor connects to) be causing this? I'm sort of doubtful of this since I swapped one over from another truck and it still did it, but I guess I shouldn't assume that that one was okay either.

The ESC module monitors the knock sensor and sends a signal to the distributor module to advance the timing or retard the timing when needed. I doubt it would cause the problem you are having.

[66gmcguy]

Originally Posted by ChevyTech

The ESC module monitors the knock sensor and sends a signal to the distributor module to advance the timing or retard the timing when needed. I doubt it would cause the problem you are having.

Well then I guess this goes back to a previous question, could there be something wrong with the PROM? The donor truck I got had been sitting for about 3 or 4 years before I got it.

[ChevyTech]

Originally Posted by 66gmcguy

Well then I guess this goes back to a previous question, could there be something wrong with the PROM? The donor truck I got had been sitting for about 3 or 4 years before I got it.

I think it is unlikely but is possible.

That ECM is so common you should be able to find something to try is even if it is not the exact same PROM.

I will be thinking about other possible causes of your trucks issues and see if I come up with some new ideas.

If you can log data and post it I would surely look it over.

[66gmcguy]

Originally Posted by ChevyTech

I think it is unlikely but is possible.

That ECM is so common you should be able to find something to try is even if it is not the exact same PROM.

I will be thinking about other possible causes of your trucks issues and see if I come up with some new ideas.

If you can log data and post it I would surely look it over.

Well, you are certainly onto something by having me disconnect the O2 sensor. That seemed to fix the cold drivability problem. It didn't run perfect still, but it was certainly a lot better.

Also, I can't say definitively, but it seemed to fix that weird misfire at a certain RPM. It may have also fixed the fluctuating idle problem, but I'm not ruling that out yet because I want to try and reset the idle speed and check the IAC counts before I determine that. I didn't have time to do that this evening but I will tomorrow.

So the next question is what the heck is wrong? The O2 sensor is clearly working. It might not be working right, but it is working as it's reading voltage. The reason I switched to a heated O2 sensor is because I needed to move it from the manifold to somewhere else because the stock O2 sensor in the manifold interfered with my oil pressure sending unit. I should note that I did have a new 1 wire O2 sensor in the manifold before, and still had the same problems as before though (apart from the cold drivability problems.) Obviously this doesn't mean the O2 sensor is not the problem, but I'm wondering if something is going on when the system goes into closed loop mode.

[ChevyTech]

With the oxygen sensor unplugged the system is not adjusting the fuel mixture. It is just using the what the PROM (chip) thinks it should need.

When the system controls the mixture it is trying to get a more precise mixture for what is needed. Some sensor data is causing problems.

A heated oxygen sensor is a more advanced and accurate sensor. The 1227747 ECM is slow and crude.

I first noticed the heated oxygen sensors being used on the S/T trucks, and this was on the OBD1 systems that were slow. They had trouble staying in closed loop and meeting emissions with a one wire sensor. I have no idea if the PROM calibration had to be changed to accommodate the switch.

Here is the next thing I want you to try.
Connect the oxygen sensor but disconnect and plug the vacuum line at the EGR valve. Then see how it runs as it warms up.

[Just call me Sean]

My oxygen sensor would go into negative voltage, making the computer think it was running lean so it would dump the maximum amount of fuel it could into the engine. Within what voltages does it operate?

[Mr_Rich]

You got all complicated at first and I didn't know how to reply to your post. Now that there are comments that are weaned out I'll tell you what I think it is since it sounds like the same symptoms I experienced when I swapped out my original 305 engine in my truck for a 350. It will run but it will definitely give you lean codes and stumble. You need the correct PROM/ Programmed PROM for the engine you are running. It sounds like you've done everything but this. That's one of the things you have to consider in this day and age of computer controlled vehicles.

[66gmcguy]

Originally Posted by Mr_Rich

You got all complicated at first and I didn't know how to reply to your post. Now that there are comments that are weaned out I'll tell you what I think it is since it sounds like the same symptoms I experienced when I swapped out my original 305 engine in my truck for a 350. It will run but it will definitely give you lean codes and stumble. You need the correct PROM/ Programmed PROM for the engine you are running. It sounds like you've done everything but this. That's one of the things you have to consider in this day and age of computer controlled vehicles.

As stated above, the PROM and the engine match.

It's worth it for me to note that I didn't just cobble the system together from junkyard parts from various different trucks. Everything (including the engine) came off the same truck. The donor truck was an 87 R20 (2wd 3/4 ton) with a manual transmission. The truck it's in now is an 82 K20 (4wd 3/4 ton) with a manual transmission, so that means that the PROM is a close match for my truck.

[66gmcguy]

Originally Posted by Just call me Sean

My oxygen sensor would go into negative voltage, making the computer think it was running lean so it would dump the maximum amount of fuel it could into the engine. Within what voltages does it operate?

Just glancing down a couple times at WinALDL while driving to my shop this morning, I saw anywhere from .004 volts (when coasting) and as high as .894 when under load. I will try and get a data log today that will hopefully give me some better data.

[Accelo]

A good friend of mine had issues with his TBI. Turned out their are two systems and they require different fuel pressures and as a result different pumps. The auto parts store sold him an incorrect pump. He changed the pump and everything went back to normal.

Would be interesting to see the results of a smoke machine test.
Cheers.