Still trying to get this thing running right!
Problem: Rough idle, backfiring through intake, eventually dies with no re-start for a while, once it dies SES light come on and I get a code 42, Electronic spark timing or by-pass fault.

When I first got the car it had codes 34 and 41, I replaced the cam sensor and magnet and MAF sensor, no change in the rough idle but it was still driveable.
One week later I start getting the backfiring and stumbling and eventual engine stall with no immediate restart.
So, I changed the plugs(they were worn to .080 gap) plug wires which were the factory originals and replaced crankshaft sensor and replaced coil pack and ICM with a junkyard unit. That seemed to fix it except for the rough idle, drove it around the neighborhood(maybe 4 miles)everyday for about three days and everything seemed fine(except it still had the rough idle)then the backfiring and stalling started again.
I inspected any exposed wiring and opened up and inspected the ECM, didn't see anything burnt.
So, before I buy a new ICM is there anything else I can check or test?
Any help is greatly appreciated!!

Can't edit my posts!! Anyway the vehicle is a 1989 Olds 98

When you swapped the ICM, did you replace it with an identical ICM (2 rows of plug wire posts, one each side) or did you swap it with a newer style Delco unit (3 individual coils, plug wire posts all in a single row)?

It may be also good to check the ICM wiring harness (the wires like to fray and break), and the ground under the ICM. Also, a heat transfer medium is recommended under the ICM.

If you replace the ICM again, I recommend switching to the newer delco version. They tend to be more reliable, and easier to service.

The ICM/Coil pack was the same, 3 coils, two rows of posts. What years were the newer version used on? The ICM grounds to its mounting plate?

Thanks!

Pulling it from a 92-95 L27 will be a direct bolt on to your application.

There is a primary ground of the underside of the bracket that can cause issues if it'* loose.

The ICM tends to transfer heat off through the bracket however.

I'll give that a try, I have a 30 day warranty on my junkyard unit lol...I'll see if I can find one tomorrow

Thanks for the info!!.

Please excuse my ignorance as this is my first experience with the 3800 but what exactly is L27?

It'* the Naturally aspirated 3800 that was used 92-95. It'* essentially the LN3 (what should be in your Olds) with some small tweaks. It'* typically identified by "Tuned Port Injection 3800" on the engine cover.

And to help clarify the code, the ICM controls spark below 400 rpm. Once it hits 400 rpm, the ECM takes over from the ICM, controlling spark timing and such.

Code 42 will typically set when the transition does not happen as expected. This can be a result of a wiring fault, ECM Fault, or ICM fault. I have also seen the code set in instances where the engine unexpectedly drops below 400rpm, such as a stumble or stall that takes a few seconds to completely die. Once the code trips, the car can immediately be restarted, however it loads default timing tables which can cause the car to run very poorly.

If the problem is intermittent, or caused by a rough stall, the key can be turned completely off for a few seconds, then back on, and the car should start fine, and the SES light should go out. The code should no longer be active (but will remain in history) and the car then should drive fine.

My 1990 Olds 88 had issues with the ICM (that appeared fairly new mind you) when I got it. I suspect one coil was bad, and the car was inherently temperamental, especially when cold. The swap to the newer style ICM corrected many issues.

My car was still plagued by other issues, however it ended up being resolved by replacing a defective battery and ECM.

An internet search for "Delco ICM Conversion" should point you in the right direction.

As Rjolly stated, the code 42 is usually set by bad wiring harness/female connection in end of harness, bad ECM, or bad ICM......

Majority of the times, I have seen it be the ICM.....however, those ECM'* were known to be bad too.....the usual wiring problem is the connector at the ECM...sometimes the female connections are loose and make a bad connection....

However, since the code 42 was not set until after you changed the ICM, leads me to think it'* the ICM....