Random high idle

[MIPS]

Started happening midway through the week. I would be coasting, at a light or just parked in the driveway and the engine idle would randomly go up a few hundred RPM and then go back down again without touching anything. Found that the solenoid side of the sol-vac was kicking in and out. I've verified it's not the headlight switch or rear defroster switch doing it so it has to be the computer. I disconnected the vacuum switches, coolant temp switch and air cleaner temp switch and it's still doing it so it isn't a faulty sensor. Unplugged and jumpered the spark control on the ignition module to verify it wasn't that freaking out and wasn't it either. Currently feeling around the harness for a rub-out but I'm not seeing anything so far and the last step will be to pull the computer and test it. Any ideas?

[AMC of Houston]

Is it the vacuum side or the electrical side of the Sol-Vac kicking up the idle?   That'll help point to a culprit.

[Taylor]

Could be the computer.
Question, does the solenoid have any voltage at idle? If it does. Does that voltage change? Could be that it is sticking?
Does the vacuum change?
Just a thought.

[Illeagle1984]

Is your A/C on?  It may bump up the idle when the clutch kicks in, then back down as it turns off for a bit.  Just covering the bases...

[MIPS]

Is it the vacuum side or the electrical side of the Sol-Vac kicking up the idle?

On further testing with the live jig, it's both. I can see the idle solenoid and the Vacuum switching solenoid operating.
I don't have AC so its just the lights and defroster. I can verify both are off, so it's the computer calling.
An unplugged vacuum switch assembly and coolant switch assembly shouldn't cause this. The Thermal Electric Switch on the air cleaner will. I can replicate the condition by shorting the pins (switch closed, air temp is cold) but as soon as I remove the jumper it settles down again. When giggling wires I found that the engine freaked out again when I lifted the air cleaner off to inspect a whistling (the manifold is probably leaking) but I also found that the grounding for the computer was about to fall apart, so I'll need to replace the grounding eyelet and clean the TES and test again.

[MIPS]

I am going to make a hypothesis.

Now that I'm trying to catch it in the act it decided to stop acting up. Go figure. Finally last night I pulled into the driveway and it was stuck in high idle so I popped the hood, jumpered the TAC switch and the idle went even higher. Unplugged the jumper and the RPM lowered to where it was originally. So no, it's not a bad TAC switch but the idle control solenoid is in fact energized, so the computer is absolutely calling for a high idle for some other reason and still responding to other sensor conditions. Went to grab tools and when my back was turned the engine stalled dead without a fight. Bumped the key and it started right back up and the idle settled down.
Was driving to work this morning (it's pretty cool out) and again, the engine stalled dead without a fight while exiting the highway. Pulled over, bumped the key and it started up. Didn't have an issue the rest of the way.

I'm suspecting the SMP ignition module I installed six months ago is part of the problem. Why do I suspect that when the computer is calling the shots here? The engine dropping dead with no warning is classic module failure and we know the aftermarket modules aren't great. I've mentioned this elsewhere but for 1982 the computer has no direct control of the TACH signal from the distributor pickup (you can't do a classic "ECM Test") so a bad computer can't kill the engine. All it can do is electronically retard the spark separately through the module and if you unplug that one wire the car will still run, it will just run with full electronic advance. I don't think the computer has the ability to retard the timing so far as to stall the engine, or at least at a high idle you would absolutely hear it fight because then you have full vacuum and mechanical advance idling at something north of 1000rpm.
The ignition module also generates the TACH pulse for the coil and the computer. If something about that signal leaving the module is weird I think the computer in some instances will kick the idle to fight low RPM.

I'll have to do some further analyzing with the ET-501. It's still possible the computer has failed since it was last tested three months ago but there's a bunch of other variables. I still have my old module and it was a working pull. It just needs new sockets because the plastics have all crumbled and you can't buy the male sockets separately.

[MIPS]

Aha. 

So I plugged the tester in, waited for the car to warm up, heard it go into low-idle closed loop and as it was checking it did its idle weirdness and immediately faulted it on the altitude control circuit.



Retested and it didn't act up again and passed, so it's everyone's favorite: an intermittent fault! \o/

Honestly I'm not sure about this part of the computer system. From what I know the altitude circuit is an extra eyelet that if you are below 4000 feet you don't attach it to a grounding point. The altitude pin on the diagnostic plug is then left floating at battery voltage. If the jumper is grounded it pulls the pin to 0v and tells the computer you are above 4000 feet, then tweaks the computer and timing. That is probably my weirdness.

Thing is, the altitude jumper wire and even the pin on the diagnostic connector is completely omitted from the schematic in the 1982 TSM. The wire going to the pin on the diagnostic connector is pink/salmon, the internet says the jumper is black with a white tracer, but the schematic says that's the computer's ground and I have a grey wire with a white tracer that shares the same crimped eyelet used to ground the system but exists on none of my schematics.



Here it is with the terminal recrimped after being found on the brink of falling off. If this was factory hardwired for high altitude we can see it's grounded here, but at the diagnostic pin it's still battery voltage, so there's a wire open somewhere that is occasionally reconnecting to ground and I have yet to find it.

The 501's troubleshooting charts say to locate wiring faults and if none are found, test the MCU. If the MCU passes it wants you to start the diagnostic over again. If it's an intermittent in the computer I don't think it hurts to just sit there and loop the test 10 or 20 times. That would HAVE to catch it.

[MIPS]

I got a full work schedule until next week so I can't pull the car apart and do any other soft of an extended diagnosis, much less find time to poke around with a multimeter. In the meantime I have ordered another ignition module (that's at least branded ACDelco this time instead of Standard Motor Products, if that means a lot) and I found my spare computer and soldered the plug back onto it so I can at least verify the worst case scenario.

If anyone has any extra suggestions leave them here and I'll look at them when time frees up.

[MIPS]

Okay so I ran through the diagnostic flowchart and verified that the high altitude jumper is not currently active, then I verified the link from the diagnostic plug to the computer plug was intact and not damaged. Sill no idea on the grey wire, but I can see for certain this isn't tied to ground and nothing about it WANTS to short to ground, so this is hardwired to never be high altitude.
Next diagnostic step was test the computer.


Tested it a few more times with both low altitude and high altitude selected, then heated it up with the hair dryer and tested it several more times. The computer is not the fault. The altitude wiring is looking less likely either. 

Started over again and went back to guessing the high idle is reactionary and the computer is responding to something I can't see. Well like I said I suspected the ignition module but even the old one acted the same way. Then I realized I had added a connector between the module and the rest of the car for an immobilizer.



That should be crimped AND soldered and fine for an amp or two, but it's overheated for some reason. I bent the spade terminal and tested again. Problem went away. It was a bad connector I had installed. Mystery solved.

[vangremlin]

Good to hear you solved the mystery.  Following this thread I thought you would make a good detective lol