1980s ECM rebuild or replacement - 1981 Fleetwood V-8-6-4

[chrisbutnut]

Where's the best place to either get a replacement ECM (ideally GM rebuilt), or is there someone reputable who rebuilds them?  I'm getting a diagnostic code 18 on my car, which is an open crank sensor.  I had my mechanic go through the troubleshooting section in the service manual, and all of the voltages are checking out, which then points to the ECM.  This is the only trouble code that I'm getting (currently), and the car seems to run fine.  It has 35,000 miles.  Someone, prior to my ownership, removed the check engine light socket.  I started to suspect this as I did more research on the code 18, and noticed that no check engine line was coming on.  I removed the dash cover, and sure enough, the socket had been removed.  I swapped over one of the other sockets with bulbs, and the check engine light now illuminates.  I'm really hoping that if I address the ECM, the problem will go away, and will not create other issues. 

Thanks.

[76Caddy]

Fellow CLC member Bruce Roe rebuilds them.

Tim

[TJ Hopland]

Bruce I think only did the analog systems that were used 76-80.  80/81 was the very first Digital Electronic Fuel Injection.  Concepts were all the same and the system lived on in the trucks through 95 but I believe these first 2 years had a bunch of 1 or 2 year only things because they were learning as they went.  Some sensors changed and I think they also changed many if not all the connectors because they found the original ones they chose were not as weather tight as they thought.

[chrisbutnut]

@76Caddy and @TJ Hopland , do you guys have Bruce's member name so I can contact him and ask?  I did a search for Bruce Roe in the membership list, but got no results.

Thanks.

[smokuspollutus]

Bruce works on the older systems. If it weren't for him and few other guys there would be 0 parts or repair capability available for the analog system. Your car uses digital fuel injection aka Throttle Body Injection which was used by GM for many years on many millions of cars. The issue comes in with the 1981 Cadillac specific aspects of the system as it was one year only, and broken down into two different versions in that year. But all in all the DFI/TBI system was miles ahead as far as reliability/serviceability when compared to the older stuff.

I have never fought this particular issue. However, in your search, note that there were two different versions of the 1981 ECM. They are not compatible so you have to get the one that's right for your car. The service manual should have whatever the breakpoint was included in it.

I have changed ECMs in a few 1981's and the "parts house" supplied ones that you put your PROMS in are a bit spotty. But you can sometimes get a good one. If I owned a 1981 I would seek out an appropriate version ECM with matching PROM's to my car and keep it on standby. The 1981 PROMs also seem a little bit spotty themselves and time has not made that better.

[Eric DeVirgilis CLC# 8621]

FI cars 1975-1979 had analog ECU which is repairable due to their design & relative simplicity. 1980 & later are a different animal which are digital and TBI injected while 1975-1979 is port injection.

I don't know if I'd want to start disturbing old wiring and connectors, potentially opening up a Pandora's box for something so trivial. It's an old car where minor defects are a part of life.

[chrisbutnut]

FI cars 1975-1979 had analog ECU which is repairable due to their design & relative simplicity. 1980 & later are a different animal which are digital and TBI injected while 1975-1979 is port injection.

I don't know if I'd want to start disturbing old wiring and connectors, potentially opening up a Pandora's box for something so trivial. It's an old car where minor defects are a part of life.

Eric, I don't disagree with you.  The "open crank sensor" fault doesn't seem to have any negative impact on how the car runs, at least not at this point.  It does illuminate the check engine light, which I'm sure is why someone previous to me just removed that light completely.  That makes me a bit uncomfortable.  I don't want to drive around with a check engine light on, though, so I guess my options are to remove the bulb again and just live with it, hoping it doesn't develop any further issues, or try to get a replacement ECM.

[smokuspollutus]

Thinking a little more about this-the "crank sensor" is really the ignition module/coil inside the distributor. Is it possible that this part changed with the intra-year revision to the engine control system, and at some point in the car's life the module was changed with a module that is incompatible?

Conversely, is it possible that there is no issue at all, but the ECM was at some point changed with an incompatible one creating a fake code?

Might want to do some digging in your service manuals and see what parts were impacted by this change-and verify the ones you have on the car will all jive.

Eric also raises a good point. Many of these parts have fragile connectors, so extreme care should be used when removing. And luck. The ECM in particular uses what are known as "edge board connections" which are really archaic computer connectors. Sometimes just unseating and reseating the connection can cause pins inside the connector to bend and not make good contact with the ECM-which can cause a host of weird issues.

[chrisbutnut]

Thanks guys.  I know my mechanic checked the voltage at the ECM, per the troubleshooting tree in the service manual.  I'll have to ask him if he unplugged anything.  Hopefully not.  I unfortunately have zero service records from before I purchased the car, so I have no idea what has been done.  I'm leaning towards just pulling the check engine bulb again and hoping for the best.  The car is going to be babied and won't be driven very often.  I can check the diagnostic codes periodically to make sure nothing else is popping up. 

[TJ Hopland]

Easily found my 80 manual hoping that would be the same but it doesn't even have that code also noted at least at a glance connectors are quite different.  Anyway finally found the 81 manual and like I thought this isn't a crank position sensor, its a cranking signal. 

It appears to be a wire that taps into the wire going from the key to the starter.  After it goes through a 56 pin connector it goes through a 3a fuse and then into the ECU. Manual doesn't really explain what that signal is doing and it being fused at 3 amps is a bit curious, that is way more amps than you would need to simply tell the computer they key is being turned. 

Code 17 is that its seeing that signal when it thinks the engine is running, presumably based on the tach signal.  18 is its not seeing it again presumably it thinks it should be seeing it because it is seeing a tach signal.  Those were not there in 80 and in 82 they got rid of 17 but 18 stayed at least through 84. I was just wondering if it was maybe a 1 year only signal that was a problem so they ditched it. Seems odd that they needed this signal but I suppose its just a double check that it should be seeing the tach signal when it sees this one. If it saw this but no tach then I suppose it would set a/the tach code what ever that is.

So the mechanic apparently traced and confirmed 12v during cranking at the orange ECM connector pin K  purple wire?  Do you trust that this person went that far?  And did they got extra deep and do drag test on the contacts?  The fact that someone in the past went pretty deep into this problem you got to wonder if they spread pins or stabbed and damaged wires before they gave up and yanked the light.

I would be looking close at that pin and wire and I would be in that area again anyway to get the part number off the ECU then searching online for that part number to see if you can find another one or two for a reasonable price.  Normally that is the best way to find ecus since they were often used on a lot of stuff so searching by car doesn't always get you results.  In this case since its 81 only the only thing you got going for you is it was the majority of 81 cars, not just a few.

I don't think the prom not being matched to the rest of the car would be an issue with this signal unless maybe its the wrong year ecu?  If it was an 80 that maybe doesn't even have this pin connected internally?  But its got the 81 prom so its still looking for that signal?  The manual I have doesn't list any numbers.                 

[TJ Hopland]

I just looked at the OBD list to see if that signal or engine speed were things you could monitor and it doesn't appear that they are. Lots of other stuff you can test and get data on but not that.  I was curious about engine speed because maybe there would be something there that is causing this code that maybe it thinks its seeing the engine turning (when its not) so its angry that it doesn't see that cranking signal?   

I wonder what would happen if you pulled that 3a fuse and then jumped a constant fused 12v signal into the ECM side of the fuse holder?   It should then see the signal all the time and trip a code 17 when the engine is running.  But will the code 18 go away?  Maybe its some sort of signal timing issue and having it constant will yield some clues?  Or will you hear the injectors firing as soon as you turn the key on because it thinks the engine is cranking?  That would then point to it getting a tach signal when it should not be.   

[TJ Hopland]

And if I was feeding it a signal at that fuse holder I would monitor the current draw to see if there are any signs of a draw.  With a 3 amp fuse I would expect there to be something there unless its a standard fuse and 3 is as small as they go?  I guess that's a possibility that they did us a favor by staying with a standard fuse vs some oddball special dealer only thing.

Also just did the conversion on what I assume is a metric wire size 0.5mm, thats 24 AWG so a pretty small wire if people have been stabbing and poking at it.   I'm too lazy to find the page again but I think before the fuse was noted as 3mm which would be closer to a #10 which would make sense for the wire that is going to the starter.   

[Eric DeVirgilis CLC# 8621]

An '80 Fleetwood would be the carburetor engine without any computer, codes or check engine light. 1980 Eldo/Seville with 6.0 liter wasn't V8-6-4 so again not relevant.

It should also be noted all 1981 Cadillacs with V8-6-4 were recalled for updates to the ECU due to issues that gave the engine such a poor reputation- issues that weren't the fault of the engine itself, but the electronics and/or the connectors. Regarding the Service Manual, it's important to recognize some of the information may have changed from the time it was originally printed due to the recall updates.

Under the circumstances of a collector car that is only going to see limited use, leaving it alone, yanking the bulb and periodically checking the codes is likely the most prudent approach.

[TJ Hopland]

At one time I remember seeing a list of most of the GM ECMs and many had info about what they were used in and sometimes other notes.  The list was related to DIY EFI stuff so it was to help people find and or identify ECMs.  Can't find it now.  The application lists I did find as I was thinking all that early stuff is pretty much just one year even for the carb stuff.  The carb ones didn't appear to be different between the brands but seemed to be one year only. 

EFI stuff appears to have stayed with Cadillac even after the other brands went EFI.  I suppose that could be to do with the true OBD feature of the Cads vs the other brands that required a scan tool if you wanted more than a blinking light. So it seems unlikely that you got an ECM from a different year.  Photos I can find it appears 80 and 81 and 82 are quite different and it didn't look like PROMs would physically interchange.  No great pics of the connectors but they did look pretty similar so maybe you could plug the wrong one into the car?   Mounting points looked slightly different so a clue could be if all the mounting holes line up?  Location of the connectors also looks to be different so if the cables don't look quite right as far as how they get to the ECM that could be another clue that something interesting things had happened in the past.

In searching one thing I forgot about these early units is they don't have a plug and socket sort of thing like what started being used in the middle 80's.  There is no connector on the circuit board, the socket with the wires attached just slides over the end of the circuit board and the contacts rub on the traces.  Not much to it and pretty much not even the slightest attempt at any sort of seal. If the contacts are like what was used on other equipment they are easily bent with meter/test light leads. They (the contacts) are one sided so you bend the little spring tab and there isn't anything left to contact the circuit board.  Many pin and blade type connectors come from both sides so you could get lucky and only bend one side. These are also fairly large like a few MM's wide so we are not dealing with tiny micro stuff like in later years.   
   

[chrisbutnut]

I was able to get a few photos of the ECM, including the part number:  1225120.  It's a remanufactured unit, so it was changed out at some point.  Hopefully it's the correct unit for the vehicle. 

@TJ Hopland  & @Eric DeVirgilis CLC# 8621, I appreciate your input! 

[TJ Hopland]

I have not looked at every page in the shop manual but didn't see any part numbers listed in the sections where I would think such numbers would be listed but they do talk about it in at least 3 different sections so I could have missed it if its there.

Looks like that locking bar in the connector is how they ID them.  Book says red, blue, and orange.  Red must be the one with the white body and the orange has the black body.  In the one photo with the top of the orange connector is that just a little dust ball on that one wire or is the insulation damaged?   Can't for sure tell the color either without it being a clearer photo or more context of the colors but the wire in question is purple and its in the 3rd cavity from the end.  The fuzz wire looks to be the top one in that row but from that angle can't tell if the top 2 are being used.   

Could be fuzz and the wrong orientation of the connector but maybe look or point the camera in different palaces to see if you can see the purple wire in the 3rd cavity from one end.   The red/white one clearly has a purple in a 3rd cavity, at least I assume that is the red plug.  If there wasn't the other plug that looked more orange I would say that one could be orange but since we are supposed to have red, blue, and orange the one with the clear purple wire has to be the red one.

[TJ Hopland]

1224760 looks like it was maybe the 'other' option.  It would be nice if any listings said what was different about them and if there was a way to look at the rest of the car and tell which one was correct.  What if it was early cars didn't have the pin for the crank signal?  Or was one the factory improved version so different part numbers just let them know that one of them was the updated one and the others needed to be updated?

Seems like I remember hearing that the factory update made em skip or hardly ever go into 6 cyl mode?  Anyone know if that was true?   Chris, does the 8-6-4 seem to work on yours?  Lots of hackary was done over the years to disable that usually with poor results when all it took was a single wire on the trans.  Besides workmanship an issue with many of the attempts (like just unplugging the solenoids) was the computer didn't know what was going on so would still change the timing and fuel assuming it had disabled the cylinders which make something that barely worked to start with much worse.  

Seems unlikely it was CA emissions sort of thing since that difference could easily be handled by a different prom.  Eldo/Seville had a different trans but then it would seem like they could just have listed E vs C car sort of thing. When this was replaced last time did they go with which ever one was available at the time?  If they were not plug compatible you would think there would be a note in the manual but then again the manual was likely printed before there was ever even the thought of a second version of the ECU. We can't go look at the 82 manual and see if there are additional notes because since it was only a limo option (which I have never seen actually happened, all the limos I have seen or seen pics of were 4100's (how terrible would that have been?)) I doubt they bothered to print all that stuff in the manual, they likely said if you got the 6.0 look at the 81 manual.  I'm not sure if I have an 82 manual, 84 (which I happen to have handy) doesn't mention anything about it.

We may never know.......

[Eric DeVirgilis CLC# 8621]

Two items I can clear up. The updates to the ECU were never designed to disable or otherwise impede cylinder deactivation. The main issue was the original ECU lacked the capacity to make the hundreds of thousands of decisions per second needed for the system to operate as intended. I believe IBM was called in to come up with better  ICs that were used to update the ICU. It's never been clear whether the existing units got new chips or were replaced altogether. I've heard both.

Regarding the factory limousines, the V8-6-4 was the only engine from 1981 to 1984. Only aftermarket stretch conversions of regular production models were ever powered by the 4100 engine. Factory limousines also had the THM400 3 speed while 4100 cars had metric OD trans since this is what was installed in junior models on which conversions were based. Commercial chassis got the tried and true 6.0 carburetor engine from 1980 to 1984 with THM400. Cadillac knew better than to install a 4100 in these professional cars which were large, heavy and subject to hard use. Aftermarket coachbuilders couldn't care less.

[chrisbutnut]

1224760 looks like it was maybe the 'other' option.  It would be nice if any listings said what was different about them and if there was a way to look at the rest of the car and tell which one was correct.  What if it was early cars didn't have the pin for the crank signal?  Or was one the factory improved version so different part numbers just let them know that one of them was the updated one and the others needed to be updated?

Seems like I remember hearing that the factory update made em skip or hardly ever go into 6 cyl mode?  Anyone know if that was true?   Chris, does the 8-6-4 seem to work on yours?  Lots of hackary was done over the years to disable that usually with poor results when all it took was a single wire on the trans.  Besides workmanship an issue with many of the attempts (like just unplugging the solenoids) was the computer didn't know what was going on so would still change the timing and fuel assuming it had disabled the cylinders which make something that barely worked to start with much worse.   

Seems unlikely it was CA emissions sort of thing since that difference could easily be handled by a different prom.  Eldo/Seville had a different trans but then it would seem like they could just have listed E vs C car sort of thing. When this was replaced last time did they go with which ever one was available at the time?  If they were not plug compatible you would think there would be a note in the manual but then again the manual was likely printed before there was ever even the thought of a second version of the ECU. We can't go look at the 82 manual and see if there are additional notes because since it was only a limo option (which I have never seen actually happened, all the limos I have seen or seen pics of were 4100's (how terrible would that have been?)) I doubt they bothered to print all that stuff in the manual, they likely said if you got the 6.0 look at the 81 manual.  I'm not sure if I have an 82 manual, 84 (which I happen to have handy) doesn't mention anything about it.

We may never know.......

The cylinder deactivation did work in my car when I bought it, but I have since disconnected the connector that attaches to the side of the transmission (the one where people say cut either the brown or black wire).  That did the trick.  I've also seen people say to disconnect the connectors on the tops of the valve covers.  When my system was operational, it did switch back and forth between 8, 6 and 4 cylinders, as it was designed to do.  I did notice a bit of a lag (compared to modern cars), and there was a bit of a vibration in 6 cylinder mode, as others have stated.  I really don't want the cylinders to be modulated, so running it full time in 8 cylinder mode is what works best for me.

I found info online that states that the 1225120 is the updated version of the ECM.  The 1224760 was the first version.  So apparently mine was updated.  It's not a California car, having spent its first 25 years or so in New York. 

The photos I posted where you can see the orange connector were with the glove box liner removed.  I put it back in after I took those photos, so I would have to take it out again to inspect those wires.  Now that I have the part number for the ECM, I've been able to find replacement units online.  Someone on eBay has one, and I found another one for a lesser cost elsewhere.  There is also a company in Nashville that's supposed to rebuild them.  I haven't taken the time to call them yet. 

After the "tune up" that I had done last week, along with re-sealing the valve covers with the recommended RTV sealant (someone had previously used cork valve cover gaskets on the car, which were leaking), the engine seems to be idling a bit high.  I talked to a local guy last week who has experience with these cars (he has an 81 Coupe DeVille and Eldorado) who said something about the throttle position sensor, which is mounted under the air cleaner.

Still trying to do more research. 

I appreciate all of the help and comments!   

[TJ Hopland]

I don't remember if there was 2 connectors on the trans or just one but you still want one wire connected somewhere for the kickdown when you floor it.  The one for the deactivation is a switch that engages when the trans is in 3rd(high) gear. 

For idle speed and and timing you have to follow the procedure / sequence in the manual exactly and use the factory specs.  If you don't it really won't run right because the computer doesn't know where its starting from and will always be trying to compensate.   Other than the proper order part of the process is putting the computer in the proper modes so its not trying to compensate as you are adjusting.  I would assume the TPS is part of that process and I think has test points just for setting.

Just about every TBI car I have ever bought or worked on was messed up sometimes to the point of not running at all and a glove box full of modules.  Every time all it took was starting at step 1 and going through the full process.   Timing was usually off.   Throttle stop was wrong.