TBI symptoms changed (new thread to reflect new problems)

[YakkoWarner]

I started a new thread because the previous one where we were trying to solve random shut-offs may no longer be relevant...

Over the weekend I had an out of town music job which meant I needed to run the 1989 Suburban out there (about 70 miles each way). After replacing the fuel pump relay which by all logic should have NOTHING to do with the random shut-offs, it ran all the way out there and all the way back without ever losing the ignition/fuel pulse. That should be a good thing (and in some ways it is), but now other seemingly unrelated problems popped in.

I had no trouble on the way out at all, just my intermittant 44-LeanRun error which is supposedly caused by a vacuum leak which cannot be detected. I think this is a separate issue from the electronic gremlins. I checked the codes once I got to the venue, only error 44.

After the show I had to make a very short trip (2 miles) for food - on the way out it ran almost perfect...it experienced one minor hesitation (almost more like a stumble) which didn't trigger any codes. On the way back, just the intermittant error 44. Loaded all my gear back in, and idled my way out of the parking area (long winding path) just fine, no codes showing.

Once on the actual road, as soon as I tried to accellerate it started bucking and audibly popping back through the air cleaner, and of course immediately the SES light came back on. I ran slowly up the shoulder knowing there was a gas station parking lot about a mile or 2 up the road where I could safely stop, and any time I tried to use more than a slight hint of throttle it would pop and buck again. Out of desperation (and remembering what was happening last week) I clicked it into neutral and switched off the key, then restarted. The restart cleared the popping and bucking - I was able to run at the speed limit from that point, but it sounded almost like there was a misfire on one of the cylinders...by the time I got to the gas station I had intended to pull off into even that had cleared up and it proceeded to run just fine the entire remaining 60-something miles home without any further issues. Once home I checked the codes again and now I have a 32-EGR error, and a 42-SparkControl error, but at no point did it ever actually shut itself off.

At this point I'm afraid to drive it anywhere. Power cycling the ignition system should not fix anything - if the module was bad or overheated, off and on wouldn't magically cool it down. A bad ground wouldn't just become good again. The coil wouldn't just start working again. The turn it off and back on again trick will stop working at some point. The only things that SHOULD be able to make it pop back through the air cleaner like that would be a bad intake valve, serious head gasket failure or incredibly over-advanced timing. None of which should change merely by power cycling.

I was unable to run the checklist from the manual for the no-spark/no-fuel issue because that never reappeared. And when I start it up at home in the driveway, it sits and idles fine for as long as I let it.

If there were any shops that still worked on these things I'd be making the appointment today - I have reached the end of what I can do at home. With only generic vague error codes as guides, there is no real way to determine what else could be happening. This feels for all the world like bad grounding problems, but everything I have read says NOT to ground the ECU itself because that can cause weird current loops. The negative cable from the battery goes right to the engine, if you're not supposed to ground the ECU to the body that presumably there must be a ground somewhere on the engine for it, but I'm 100% stumped on where.

I'm starting to think it is either haunted, or somwhere in the bowels of the black boxes there is a copy of Windows 3.1 running which needs to be rebooted every couple hours...

 

[gmbellew]

code 42 is fairly limited, in that it is an issue with the ICM, either in the EST (white) or Bypass (tan/black) wires. they went open or saw unexpected voltage spikes. I'd be looking hard at the ICM connector itself, the wire from the external coil to the cap for signs of spark leaving the wire, and inspect the distributor cap for cracks that could be letting spark out.

potentially you could have an intermittent rub through on the 440 circuit ( fuel pump and ECM power). I had a rub through that was causing similar issues and also threw strange and unrelated codes. sometimes it would pop ECMB fuse, sometimes not. locations to check are by the cruise control bracket and MAP sensor bracket for where I have had rub through.

your Lean O2 is probably not related. but it is likely an issue with O2 sensor itself or the wiring to the O2 sensor if the fuel pressure is good.

 

[YakkoWarner]

I do have another unused new cap that I can swap on in place of the one that came with the distributor. Since it was well after dark when I got home I was able to open the engine compartment and look for any arcing on the high-voltage plug/coil wires. Did not see any but that doesn't always mean its not there.

Rub thru near the MAP bracket is a real possibility, that could explain why the sudden shut-off disappeared as a result of unplugging/reconnecting the relay connectors that are right behind that. Causing the computer to have a momentary voltage drop would certainly make it act lobotimized. All my cruise control stuff was removed by a previous owner (at least the engine compartment side - everything under the dash is still present) and the vacuum line that went to the actuator has been plugged. As a result, I'm not sure exactly where the bracket would have been. ECM-B fuse has not popped, but I replaced it with a new one in case the very old original one had some sort of hairline break.

An out of the box thought - would there be any benefit to wrapping the ICM wires in shielding foil to eliminate any stray inductions (if that is what is causing the 42 error)?

 

[WP29P4A]

"A bad ground wouldn't just become good again." Oh yes it can. A ground that is disconnected or cut can not reconnect it self, but a bad ground due to corrosion or a lose connection can very well change from good to bad and several steps in between, from heat and cool cycles, humidity and load on the connection, as well as vibration.

No ground and a bad ground are not always the same thing. The hardest grounding issues to find are the ones that are between a good ground and no ground. When you are between those two points you are dealing with resistance.

Where is your TBI ground and how does it look? What is it attached to? Keep in mind, electrical is a TWO sided deal. If your grounds are not solid your electrical will be intermittent if it works at all. One of the common things I have seen in the 45 plus years I have been trouble shooting electrical problems is, people spend plenty of time making the + side of their electrical system work and look great and totally neglect the - side of the same system, as if it's less important.

 

[YakkoWarner]

I think you are correct on this one - and after dealing with British electricals for years I have learned to always question the grounds and never directly trust the body to be a good ground (or a trailer frame, especially a boat trailer).

I'm out of my territory so to speak on this one though...I don't even KNOW where the grounds are supposed to be. It looks like there are multiple ones based on the wiring diagrams, but I have no idea where they are. My initial instinct was to solidly ground the ECU to the body, or run a dedicated ground wire thru the firewall to the engine. Everything I have read says specifically not to do this, apparently GM left the ECU isolated from the body for a reason involving stray loop currents. My statement about a bad ground not becoming good was strictly in the context of it magically restoring itself in that 2-5 second window of turning the key off and on again. I do realize that condition and environment can certainly make for inconsistant connections.

The negative side of the battery goes straight to the engine, there are also pigtails from the battery negative to the front core support. I added a pigtail from the coil bracket stud on the intake directly to the firewall as well because I used to get very consistant error 42 triggered, until yesterday I hadn't seen an error 42 in over a year so I assume that extra pigtail helped.

 

[WP29P4A]

On my son's truck the TBI ground was on the water neck. We replaced the connector on the end of the wire and cleaned the grounding location, and it eliminated all of the intermittent TBI issues. (Not all the issues, just the intermittent ones)

 

[gmbellew]

I think the 2 engine grounds for the ECM are the back of the block on the passenger side and by the thermostat and CTS in the front of the intake manifold

 

[gmbellew]

I do have another unused new cap that I can swap on in place of the one that came with the distributor. Since it was well after dark when I got home I was able to open the engine compartment and look for any arcing on the high-voltage plug/coil wires. Did not see any but that doesn't always mean its not there.

Rub thru near the MAP bracket is a real possibility, that could explain why the sudden shut-off disappeared as a result of unplugging/reconnecting the relay connectors that are right behind that. Causing the computer to have a momentary voltage drop would certainly make it act lobotimized. All my cruise control stuff was removed by a previous owner (at least the engine compartment side - everything under the dash is still present) and the vacuum line that went to the actuator has been plugged. As a result, I'm not sure exactly where the bracket would have been. ECM-B fuse has not popped, but I replaced it with a new one in case the very old original one had some sort of hairline break.

An out of the box thought - would there be any benefit to wrapping the ICM wires in shielding foil to eliminate any stray inductions (if that is what is causing the 42 error)?

I'd let the engine get up to temp, and at idle, start wiggling and manipulating the harness and see if you can reproduce the issue. If not, I'd spray some light water mist from a bottle around the cap and ICM and see if you get any sparks.
If that looks good, I'd be testing for continuity on the ICM EST amd ByPass Circuits wires from the connector to the ECM.

 

[YakkoWarner]

I assume doing this experiment will require disconnecting the 4 pin module connector at the distributor AND the 2 very large connectors on the ECU itself? It will probably be a week or so before I can dig into this but I am certainly saving these suggestions and information to use the next time I can get into this thing. On a different (but related) topic - what are other people doing to get access to the distributor/coil area at the back of the engine? I normally end up building a temporary scaffold using 2x4s and planks across the fenders or from the cowl vent to the core support (using towels or rags to avoid scratching up the paint) because I am short and seemingly have little T-Rex arms compared to whoever assembled these things to begin with.

 

[gmbellew]

I assume doing this experiment will require disconnecting the 4 pin module connector at the distributor AND the 2 very large connectors on the ECU itself? It will probably be a week or so before I can dig into this but I am certainly saving these suggestions and information to use the next time I can get into this thing. On a different (but related) topic - what are other people doing to get access to the distributor/coil area at the back of the engine? I normally end up building a temporary scaffold using 2x4s and planks across the fenders or from the cowl vent to the core support (using towels or rags to avoid scratching up the paint) because I am short and seemingly have little T-Rex arms compared to whoever assembled these things to begin with.

Working in that area sucks. A top creeper would be ideal. But I usually climb in and put my left foot on the passenger side frame rail and right foot back by the core support and lean in. Most of my weight is supported by the left foot. It's not comfy, but gets the job done.

 

[YakkoWarner]

I like those top creepers - my friend Dave had one but I didn't try to buy it from the estate. It wouldn't have helped me all that much since I only have a mud/dirt driveway (no paved parking area or actual garage) to work on, but on a smooth hard surface it was ideal. I may try your passenger side frame rail method if I can wedge myself in there - would be quicker than finding lumber.

 

[chairmnofthboard]

My 87 would buck and die at hwy speeds (cruising). It would set code 34 (MAP). I replaced the MAP and EGR with an ACdelco unit and the problem went away. I would also get a random stall around town and I replaced the ICM with a ACdelco (D1943A) and a ACdelco pickup coil (D1987) and that solved the problem.