ecm conversion, 1986 to 1989

I'm curious about the feasibility of converting the 1986 3.8 vin 3 computer system to one as found on the gen 1 3800's. I am at the end of the rope with the ecm on my LeSabre, and just happen to have the wiring harness, sensors, engine that shouldn't be run (motor spun bearings), and most everything else from a 1989 Electra PA. The car was bought in rough shape as a parts car for my running 89 Electra PA, but mostly to have an extra tranny for $150.

I am capable of many things mechanical and electrical (swapped and fully integrated a '91 3800 vin c and a '87 700R4 trans into a '82 GP - runs and drives, but not in snow as it's rwd)

I have arc welder, mig welder, gas welder, spare parts. I have absolutely no desire to put a 3800 in my '86 LeSabre.

Purpose of the project is to eliminate the '86 electronics that are not cooperating with me to something more useful with a 8192 baud serial interface for better control of the engine, and possibly increasing power or fuel economy (or in my case both). The car used to get 31-32mpg without question; rans smooth and accelerated 0-60 in 7.4 seconds flat. It now gets 21mpg reguardless of driving habits or speeds or city/freeway driving conditions, though swapping from a 440-t4 to a 4t60 tranny results in funny shift points for a motor that redlines about 500rpm higher than mine. ECM shows no faults and serial data is withing acceptable parameters, yet brand new rebuilt engine vibrates and exhaust spits (differently than miss-fire) and has no power at mid to upper RPM range, the same symptoms present before rebuild. Cold air intake and dual exhuast determined to not be effecting power or mileage. Swapping the 1-4, 2-5, 3-6 injectors does not effect power or mileage. I have over 6,000 miles of computer serial data, and the only concern is abnormal spark retard that I cannot eliminate at all. Every sensor and solenoid, controllers and modules have all been replaced and several ecm/prom combination tested; results are conclusive that I can't fix the spark retard problem with the system... but I digress

Any information on fuel maps, spark adv. tables, or anything else like that for both would be helpful to my plight. I can read/burn PROMs if needed, though would like to avoid that if at all possible. Also, any thoughts can be redirected to the thread I started in the LeSabre forum (reguarding poor fuel economy and poor performace) for possibly solving the root of the issue (thread that more accurately and fully describes the current problems with the computer, though the LeSabre forum seems to be inaccessable at the moment).

The 3800 is balanced differently compared to the 3.8, uses a tuned port injection setup (or sfi, depends on the year) compared to multi-port, blah blah blah blah. So it's a crapshoot. By the time you are done, you WILL basically have a 3800 in your LaSabre.

Your missing something big here. HUGE. Unless you haven't mentioned it, and if that's the case please do.

Rule #1: Electronics always come in second when diagnosing problems if there is no MIL lit. Basic mechanical system are first.

What's your compression? How's your fuel pressure? How's the air filter? Coolant Level? What color is the oil? Spark plug/wire condition? Tire pressure?

You have hours of ecm data, so is the TCC functioning properly? Those shift points are adjustable, and probably have a lot to do with your issues. Get to a tranny shop and have them tweaked back to where they belong.

well, it may be controlled to run like the 3800, but there is a significant difference between the block of the 3.8 and the 3800: The block is not as tall, so there is a smaller volume in the chamber. Our 3.8 would push 32mpg freeway consistantly, where the 3800's we've had never broke 28 in the same driving conditions.
Balancing is the only real issue; and with the help of the Service Manuals I can modify the balancer to have the correct sensor interupts.

Answer to rule #1: The problem was present prior to a bearing spinning. The problem appears to have originated at the time when the original nylon timing gear on the camshaft stripped out and the chain came off. In the process, the original computer and ICM were replaced trying to troubleshoot the problem (hard to start, but ran perfectly once running - 32 mpg - until it stopped one day and there was plastic in the oil pan). I was not the "troubleshooting mechanic" at the time (still in gradeschool...) but there is a long story involved with never getting the new computer to ever work the way the original one did (original PROM in new computer). *edit* The original computer was turned in for a core charge prior to recording any numbers on it.
After it spun a bearing when I first got the car, I rebuilt the engine from the ground up. I went by the book (chassis service manuals), checked, double checked the engine mechanical functionality. I also rebuilt the trans because it lost 3rd gear. Still no go, and after hours of fixing intermittant trouble codes from sitting for 3 years it still runs like crap. I had used a '87 engine temporarily and even that would only get 25mpg. My friend has an '87 with the same engine and he averages 31 mpg on about 190k miles. I suspect that it is possible that a different computer model could have been the one originally installed in my car at the factory, or that the cam sensor is 180 degrees off and it's screwing something up. I can make an audio file of the exhuast and send it to you if you think it could supply comic relief. Injectors are firing sequentially but spark retard is always present and is directly related to rpm (as high as 35 degrees retard at 5000 RPM)
I've just learned that the '86-'87 Riviera vin 3 has a 8192 baud computer, so I may try to see if that computer has the same connectors on the back.

ok, that's odd. spark retard is proportional to rpm?
You know where I normally see that? On (turbo)supercharged engines, spark timing is retarded as boost (and therefore rpm) increases. Yes, your ecm is acting as if it hit it's head and woke up thinking it was in a grand national or t-type.
Your Thoughts?