GM TBI on a 1966 283 Engine - Getting a Code 42

[1project2many]

The ignition system is designed so the ICM operates independently of the ECM until the ECM is ready to take control of spark timing. The ECM signals it is ready by applying 5V on the tan and black "Bypass" wire connected to the module. Some of the vehicles equipped with '7747 ecm's had instructions to set ignition timing by jumpering two pins in a connector under the dash. Others used instructions that said the Bypass wire should be disconnected under the hood. Problems can occur if you use the wrong set of instructions as the ignition timing can be off significantly.

If the ignition timing is correct it's possible to force the ICM to ignore the ECM. This can be an important step in diagnosing a spark problem. If the ECM is disconnected and the coil stops producing spark it's obvious the ECM is not the culprit. Locate the tan/black Bypass wire to the ecm and disconnect the wire to keep the ECM out of the picture.

Even though the engine is not equipped with a knock sensor the ECM still uses knock counts. The value in the knock counter will be different each time the engine is started but should not change during the time the engine is running. As long as they do not change while the engine is running the ECM will think there is no knock.

[nvestysly]

The ignition system is designed so the ICM operates independently of the ECM until the ECM is ready to take control of spark timing. The ECM signals it is ready by applying 5V on the tan and black "Bypass" wire connected to the module. Some of the vehicles equipped with '7747 ecm's had instructions to set ignition timing by jumpering two pins in a connector under the dash. Others used instructions that said the Bypass wire should be disconnected under the hood. Problems can occur if you use the wrong set of instructions as the ignition timing can be off significantly.

If the ignition timing is correct it's possible to force the ICM to ignore the ECM. This can be an important step in diagnosing a spark problem. If the ECM is disconnected and the coil stops producing spark it's obvious the ECM is not the culprit. Locate the tan/black Bypass wire to the ecm and disconnect the wire to keep the ECM out of the picture.

Even though the engine is not equipped with a knock sensor the ECM still uses knock counts. The value in the knock counter will be different each time the engine is started but should not change during the time the engine is running. As long as they do not change while the engine is running the ECM will think there is no knock.

Okay, I know this sounds simple to many on this forum but let me restate what I think you're saying to see if I understand the process...

My instructions indicate to set the timing with the bypass wire disconnected. My factory timing spec for this engine is 4 degrees. I disconnect the bypass wire and set the timing at 4. That creates a code in the ECM and the CEL is on. I have to clear that code by disconnecting power to the ECM. I reconnect the bypass wire. Start the engine, runs fine. The timing is now advanced quite a bit at idle - maybe 12, 14, 16 degrees. I don't know for sure but it's certainly advanced and the engine is running fine. That means the ECM is controlling the advance, right?

So in the past, when the engine stumbled, the CEL came on and I get a code 42. The timing reverts back to 4 degrees - I've measured it in this condition and it's definitely back to 4 degrees. In some ways that's the same as disconnecting the bypass wire. The engine still runs, I guess with no advance, so it's a bit sluggish and I can limp home in that mode. However, when this happens the code 42 doesn't seem to be a "hard" code. As soon as I turn off the ignition and restart the engine everything is fine. The CEL is off, the spark advance is back to where it should be and all is well until the next time the engine stumbles.

I'm probably going around in circles here but I don't understand what the difference is between 1) physically disconnecting the bypass wire which sets a hard code 42 and the CEL remains on until I clear the codes compared to 2) the random phenomenon that is occurring in which it appears the timing goes back to limp home mode, indicates code 42, but doesn't set a hard code 42.

I haven't been able to idle (or drive) the car very much since I replaced the ICM. I'm now preoccupied with the fact that the engine stalls every few minutes. I've been looking into that and may have more information tomorrow.

Next time I operate the engine I'll look at the Knock Counts. I think this value changes while the engine is running but I know for sure there is no knock sensor. I'll report more on this tomorrow.

Thanks again for the replies. I'm learning as I go. As I said in my introduction thread I don't feel I know enough to be a gearhead, just enough to be a dangerous shade tree mechanic!

[1project2many]

I would view this situation as a box of unknown problems. The computer tune may or may not be right for the engine. The wiring harness may or may not have a problem. The ignition module may or may not have a problem. The distributor may or may not be working correctly. System voltage may or may not be causing a problem. Or the Gods of Mechanical Things may or may not be entertaining themselves by altering the laws of physics and logic around the car whenever it's in use. The goal, to me, would be to solve for unknowns while watching out for wrenches falling horizontally and bolts of electricity randomly leaping from the cat.

I'm probably going around in circles here but I don't understand what the difference is between 1) physically disconnecting the bypass wire which sets a hard code 42 and the CEL remains on until I clear the codes compared to 2) the random phenomenon that is occurring in which it appears the timing goes back to limp home mode, indicates code 42, but doesn't set a hard code 42.

I think you've got the basic operation down. Your observation of the difference is a good one. It is not typical behavior for the ecm. Typically if the ecm sees a problem on the bypass line it will store code 42. However, there are certain conditions where the ecm will determine there is a problem currently but it will not set the code. Those problems can be very difficult to diagnose. They are often caused by voltage fluctuations on the bypass or EST line and technicians from the early days of GM EFI all have a story or two about trying to cure an intermittent code 42.

You have also reported " I try starting a few more times and ultimately end up putting the accelerator pedal to floor and hope like he!! it will start." Did the engine start while putting the accelerator pedal to the floor and cranking? Or did you need to use a different method? If it started while putting foot to floor the engine may have flooded during cranking or just before stalling, possibly an indication that more tuning is needed. Which may or may not be part of the stalling problem.

You also said "So in the past, when the engine stumbled, the CEL came on and I get a code 42. The timing reverts back to 4 degrees - I've measured it in this condition and it's definitely back to 4 degrees. " This is an excellent diagnostic step. This is also correct behavior for the ecm. When the ecm senses an incorrect voltage on the bypass line it will flag a problem with the check engine light, send up a code 42, and stop attempting to control ignition timing. Timing should revert to base timing, in your case 4 degrees. This is also a good sign, at least as far as discounting the theory that the gods are playing games with your car.

Please tell me if this list covers all the symptoms:
1) Check engine light / "soft" code 42 occurs under various conditions
2) Engine stalls very frequently after ICM replacement, primarily at low speeds. Restart can be difficult.
3) ECM will, at times, command increasing IAC counts causing very high engine speed

Is this list complete? Did I miss anything?

[nvestysly]

Yes, that captures the basic problems I’ve experienced. The past six weeks are a bit of a blurr so I sometimes have difficulty recalling which problem retates to which period of time and how that may or may not relate to what’s currently happening.

The latest issue is certainly the frequent stalling/dying for no apparent reason and it’s not setting a code.

[nvestysly]

Please tell me if this list covers all the symptoms:
1) Check engine light / "soft" code 42 occurs under various conditions
2) Engine stalls very frequently after ICM replacement, primarily at low speeds. Restart can be difficult.
3) ECM will, at times, command increasing IAC counts causing very high engine speed

Is this list complete? Did I miss anything?

I'm back at home now with access to a regular keyboard so I can type more easily. The earlier post was from a smart phone.

So I'll elaborate on item 2) Engine stalls very frequently after ICM replacement, primarily at low speeds. Restart can be difficult.

I recall now that stalling/dying occurred with the previous ICM but it was much more infrequent. It was at times difficult to start with the previous ICM as well as the replacement ICM. The situation does seem to be worse (increased frequency) with the new ICM but my guess is that's a coincidence. My guess is that whatever is causing the engine to stall/die is some other issue but maybe affected when I replaced the ICM.

I just ran the engine for 5 or 6 minutes and it died again. Upon trying to restart I was able to check for a spark at one of the plug wires where it attaches to the distributor cap. No spark. I tried restarting 4 or 5 times just a few seconds each time. No spark. Finally, after several more attempts there was spark and the engine started immediately. I did not touch the accelerator pedal. My previous comments related to pushing the pedal to the floor probably have nothing to do with whether the engine was going to start. I think I was simply reaching a level of frustration and the only thing I could think to do was push the pedal to the floor. I'm beginning to wonder (recall?) if similar things hadn't happened back in the carburetor days I just don't have a clear memory of those situations as the car has been in the garage, mostly unused, for nearly a year until the TBI system was installed recently.

Yesterday, at one point when I was troubleshooting I tugged/pushed/ pulled on every wire and connector I could find. The two connectors at the distributor, the IAC, TPS, coolant, ECM connectors - nothing I did made the engine die. So I'm increasingly believing that it's not caused by a loose wire. At least not a loose wire or connection related to the TBI install.

Many years ago the car would die but all electrical circuits would die with it. The horn, lights, radio, you name, all dead. I replaced the under hood wire harness and that seemed to fix the problem. Perhaps something like that is happening again. I'll have to do some research on which wires go where and which items are in the ignition "loop."

This thing is driving me crazy!

ON EDIT: The knock counter on the 1320 Electronics software is changing while the engine is running. The counter started at zero when I started the car but changes wildly (sort of like the O2 numbers change wildly) from 50 - 200 - 75 - 275 - 40 - 195... with every data record.

[1project2many]

I believe focusing on the stalling / hard start might be a good place to enter into this fight. The suggestion to disable the ecm controlled spark advance is a diagnostic step. This completely eliminates ecm control of the timing. If there is no spark with the bypass connector disconnected then the distributor, coil, and wiring must be checked. Alternatively if the engine stalls while the ecm is controlling timing you can disconnect the bypass connector to see if the engine restarts.

Other tests if there's no spark could include checking resistance of pickup coil and watching the laptop to see if RPM is showing greater than zero while cranking. I've just gone through this today with a truck that only sees daylight a handful of times at best.

[jim_in_dorris]

I am going to second 1project on this. Is actually sounds like a bad pickup coil in the distributor. I had that happen on my s10. It would die, then after sitting it would usually restart. On all tbi engines my mechanic always replaces the distributor on nostarts.

[nvestysly]

Thanks for the comments. This approach seems reasonable and since the stalling/dying is the overriding factor at this point I'll be looking into this. A couple of questions (which will demonstrate my lack of knowledge and understanding of the components involved):

1) How does one check the resistance of the pick up coil. Yes, I have a multi-meter. But what exactly is the pickup coil? I presume it's part of the ICM. Which pins do I check? I presume the resistance should be low but what value?

2) If the pickup coil is part of the ICM then the old unit had the same problem but did it much less frequently. The replacement ICM is stalling/dying every few minutes and doesn't specifically have to be "hot" as it has done it within 5 minutes of starting the engine. Do I need yet another ICM? If so, a particular brand?

I'll report more later today.

Thanks again.

[nvestysly]

I drove the vehicle for the first time in several weeks. The ECM remains temporarily moved away from any moisture that might drip from the air conditioning ducts. In total I drove about 50 miles today. I did not use the A/C today because the outside temperature & humidity were quite pleasant.

About 20 miles into the stop & go traffic and errands I was running I was idling at a traffic light. The car surged and I could hear the IAC counts must have increased dramatically because the sucking sound under the hood. Fortunately my foot was firmly on the brake pedal so the car didn't move much before I stepped even harder on the brake to keep it from lunging forward and hitting the car in front of me. As I said, this thing has a mind of it's own.

As I was getting close to the end of my 50 mile round trip the check engine light came on. Arggh. The engine became sluggish because it was back to base timing without any advance. I knew from past experience that if I turned off the ignition and restarted the engine the CEL would turn off and the timing would be back to normal. So when I was coasting down a hill I did just that - turned off the key for a moment and then turned it back on - and the timing went back to normal.

So here's my latest question/issue:

The idle speed on the engine is not consistent. When I go through the entire process of setting the throttle plates, the engine idle, TPS voltage, the IAC closed, and get it tuned in with the engine idling and in gear (I have a powerglide) it seems to be fine for a few minutes. If I rev the engine and let it come back to idle it doesn't return to the idle speed that I set previously. Sometimes it's higher and sometimes it's lower RPM. I'm thinking this variation in idle speed and corresponding variation in TPS voltage might be causing a problem. In particular if the idle goes a little low it seems like the ECM may try to compensate and open the IAC to bring the idle up to the set point. But this may not be the right TPS voltage because the throttle plates aren't where their supposed to be. Okay, I'm going around in circles again but this dang thing has me baffled. Why does it run fine for a period of time then go a bit crazy?

Some helpful information on the position of the throttle plates - It appears the throttle plate screw will not control the actual position of the butterflies on a consistent basis. One time I set the parameters and the next time I drive the car those parameters are off kilter. It's as if the throttle plates are not returning to the same position each time. There doesn't seem to be any slop/wiggle in the throttle shaft. What should I look at to investigate this further?

1project2many indicated I should try to follow the stalling/hard start issue and I'd like to do that but it seems to have gone away for now. I plan to drive the vehicle several times in the upcoming week so I'll keep an eye on that issue too.

[jim_in_dorris]

I drove the vehicle for the first time in several weeks. The ECM remains temporarily moved away from any moisture that might drip from the air conditioning ducts. In total I drove about 50 miles today. I did not use the A/C today because the outside temperature & humidity were quite pleasant.

About 20 miles into the stop & go traffic and errands I was running I was idling at a traffic light. The car surged and I could hear the IAC counts must have increased dramatically because the sucking sound under the hood. Fortunately my foot was firmly on the brake pedal so the car didn't move much before I stepped even harder on the brake to keep it from lunging forward and hitting the car in front of me. As I said, this thing has a mind of it's own.

As I was getting close to the end of my 50 mile round trip the check engine light came on. Arggh. The engine became sluggish because it was back to base timing without any advance. I knew from past experience that if I turned off the ignition and restarted the engine the CEL would turn off and the timing would be back to normal. So when I was coasting down a hill I did just that - turned off the key for a moment and then turned it back on - and the timing went back to normal.

So here's my latest question/issue:

The idle speed on the engine is not consistent. When I go through the entire process of setting the throttle plates, the engine idle, TPS voltage, the IAC closed, and get it tuned in with the engine idling and in gear (I have a powerglide) it seems to be fine for a few minutes. If I rev the engine and let it come back to idle it doesn't return to the idle speed that I set previously. Sometimes it's higher and sometimes it's lower RPM. I'm thinking this variation in idle speed and corresponding variation in TPS voltage might be causing a problem. In particular if the idle goes a little low it seems like the ECM may try to compensate and open the IAC to bring the idle up to the set point. But this may not be the right TPS voltage because the throttle plates aren't where their supposed to be. Okay, I'm going around in circles again but this dang thing has me baffled. Why does it run fine for a period of time then go a bit crazy?

Some helpful information on the position of the throttle plates - It appears the throttle plate screw will not control the actual position of the butterflies on a consistent basis. One time I set the parameters and the next time I drive the car those parameters are off kilter. It's as if the throttle plates are not returning to the same position each time. There doesn't seem to be any slop/wiggle in the throttle shaft. What should I look at to investigate this further?

1project2many indicated I should try to follow the stalling/hard start issue and I'd like to do that but it seems to have gone away for now. I plan to drive the vehicle several times in the upcoming week so I'll keep an eye on that issue too.

Something to try is to disconnect the throttle cable and manually flick the throttle to verify it is returning to the same place every time. If it doesn't, your throttle body is binding. If it looks good, start thinking about your throttle cable. It should be parallel to the manifold and at about the same height as the throttle connector on the throttle body. With it disconnected, tug on it to see if it feels sticky.