Old Car computer (ECM) may be faulty

[noquacks]

People,

1987 Sunbird GT non turbo, replaced O2 sensor 2 times and still the computer emits a code 45 "rich exhaust or bad O2 sensor". I replaced the pressure regulator, tested the temp sensor and throttle positioning sensor, fuel seems to flow unimpeded into the throttle body from the injector, nice cone shaped spray, even when throttle is manually bumped to accelerate.

Because my car is so old, it has some water leakage when it rains, and I just found out that the ECM unit at the base/front of the car interior was wet. Maybe this could be the problem. I have 2 pictures below. 2-3 contacts/blades seem very dirty! Can this be my problem? Can I clean with rubbing alcohol?

Also note next to the worst looking terminals are 4 (resistors??) that seem crudded up. Advice appreciated.

 

[kellys_eye]

The crud doesn't help. If you have a connection diagram you could find out if those pins are the actual pins connecting to the O2 sensor circuit and thus confirm your suspicions.

But a small fibreglass pen could clean that up in moments.

 

[Martaine2005]

The 4 diodes definitely could have problems or even a short. Plus a short to the capacitor to the left of the 4 diodes.
Absolutely clean it first with IPA (isopropyl alcohol) and the fibre glass pens are great for tracks and connections as suggested above.

Martin

 

[noquacks]

Thanks. Are "tracks" marks? oxidation? just dirt? And where are the tracks on the circuit board? On the pins, like the worst 2 on the right?

 

[kellys_eye]

The 'tracks' are actual lines of copper material that join the individual components to form the circuit. Oxidation occurs when you get an electro-chemical reaction from condensation (or other contaminant) on any part of the circuitry where there might be a potential difference (voltage) between two parts or tracks. The corrosion, if left unattended, can cause a break in the track or even destruction of the leads going to the components.

Additionally, any contamination can introduce spurious voltages, cause a loss of signal or even create a short circuit depending on its severity.

As mentioned above, dissolve the 'crud' (the whiteish stuff) using IPA (alcohol) and clean the edge connector LIGHTLY using the fibre brush so the shiny solder track reappears. If you go too far you'll expose the actual copper track and that will rapidly deteriorate thereafter.

Once cleaned, refit and test to see if the fault has gone. It should have. Car ECU's are remarkably resilient and you've caught to potential future issues by attending to it now. I very much doubt there are any damaged components.

 

[Bluejets]

But a small fibreglass pen could clean that up in moments.

Thing is, that may help with what is visible in the photos but............
There is a mating receptacle which would be almost impossible to clean with that method and both are required for a connections.

 

[H2814D]

Take a very clear picture of the other side of that board and let us look at the solder joints. The GM/Delco circuit boards from that period were notorious for having solder joints go bad after time. A "cold solder joint" is a common term for it, if you want to look it up. On your board the most common places for them to occur would be any component that is prone to heating when energized. The multiple heat cycles (heating and cooling) will cause expansion and contraction of the joint, which leads to microscopic cracks around the electronic component pin.

If you can examine the board under a magnifying glass, you may find joints that look like these (see pics). They are hard to see with the naked eye sometimes, unless you know what you are looking for. The components that heat up will often be attached to heat sinks. Look there first. Next would be any power supply inputs. To repair them, add flux (optional, but preferred) to the pin, reheat (reflow is the actual term) the solder connection to melt the solder, add a little bit more solder to that joint, and let it cool back down. That should correct the cold solder joint problem.

Not guaranteed that that is the reason for your code, but since you have the board out, it needs to be checked.

And I believe the four components you are identifying as "resistors" are actually diodes. The test procedures are entirely different, in case you didn't know that.

 

[kellys_eye]

An aside.....

you replaced ONE of the O2 sensors. Most cars have two. Did you replace the other?

 

[pete.cal]

Does it have a Mass Airflow Sensor (MAS)? I once had a problem like you describe and found a blade of grass against the egg crate in the MAS. It caused the computer to misread the amount of air entering the engine and so it couldn't balance things out.

 

[H2814D]

Another cause on a car that old is a failed/failing/clogged catalytic converter.

 

[noquacks]

An aside.....

you replaced ONE of the O2 sensors. Most cars have two. Did you replace the other?

My car has only one. Old cars this is typical.

 

[noquacks]

Does it have a Mass Airflow Sensor (MAS)? I once had a problem like you describe and found a blade of grass against the egg crate in the MAS. It caused the computer to misread the amount of air entering the engine and so it couldn't balance things out.

It has a MAP Manifold absolute pressure), not sure what the difference is to a MAS. . No engine code emitted for that sensor. But thanks for the idea. I am now looking at everything. Just got a new TPS today, even though the current one checked out OK.

 

[noquacks]

Take a very clear picture of the other side of that board and let us look at the solder joints. The GM/Delco circuit boards from that period were notorious for having solder joints go bad after time. A "cold solder joint" is a common term for it, if you want to look it up. On your board the most common places for them to occur would be any component that is prone to heating when energized. The multiple heat cycles (heating and cooling) will cause expansion and contraction of the joint, which leads to microscopic cracks around the electronic component pin.

If you can examine the board under a magnifying glass, you may find joints that look like these (see pics). They are hard to see with the naked eye sometimes, unless you know what you are looking for. The components that heat up will often be attached to heat sinks. Look there first. Next would be any power supply inputs. To repair them, add flux (optional, but preferred) to the pin, reheat (reflow is the actual term) the solder connection to melt the solder, add a little bit more solder to that joint, and let it cool back down. That should correct the cold solder joint problem.

Not guaranteed that that is the reason for your code, but since you have the board out, it needs to be checked.

And I believe the four components you are identifying as "resistors" are actually diodes. The test procedures are entirely different, in case you didn't know that.

Will send a photo probably Saturday. Thanks. Good info.

 

[noquacks]

Heres the picture of the back! Thanks!

 

[H2814D]

View attachment 62421

Heres the picture of the back! Thanks!

Thanks. As expected, there are a ton of soldered pins on the back of that board. With the magnification limitations I can accomplish on my end, there isn't enough clarity and definition to adequately examine the board for what I am looking for, so it looks like you will have to do it.

It may help if you clean the bottom of that board with alcohol before you start. Start from one end of the board and check each solder joint for failure points around each pin like in the pics I supplied. Note that the pics I supplied for examples were magnified quite a bit, so that should let you know how difficult it may be to find them without magnification on your end. If there are failed solder joints, they probably won't be found on the smaller soldered pins. Like I said, the components attached to heat sinks and power supply-in pins are the usual suspects. They will typically be the larger soldered joints. If you find any, try and focus the camera on that particular area and take more and clear pics before you attempt repair. If you can locate them and it fixes the issue, there is value to others by coming back here and posting your results of the location(s) of your bad solder joint and your success, if any, in the repair.

Additionally, if you want to try again with smaller quadrant pictures of the board and post them here, that may help us as well. But just so you know, that looks just like the type of GM/Delco boards I have experienced the very same issues with before. I've seen it on windshield wiper motor boards. Cruise control boards. Anti-Lock Brake Control Boards, stereo boards and others.

 

[kellys_eye]

Cold solder joints are rare on ECUs - my experience is that they are much more reliable than most manufactured boards as they are designed to work in harsh environments. A quick 'scan' of the image provided shows no particular issues that I can see and trying to check EVERY soldered joint is asking a bit much.

The single edge connector pad at the bottom left of the picture is VERY suspect though and needs attention - it corresponds the other 'dirty' connections on the other side of that board PLUS the contacts in the socket the board plugs in to (per @Bluejets post above).

Fix those edge connections and you'll probably find everything else works.

Do you have a schematic that might indicate what those connections are for?

 

[noquacks]

Good to know, H28 and Kelly. I have not been able to get a schematic even with the official SHOP manual from GM. But I am happy to learn that solder joints are rare on ECU's, which mine is. Note that I am looking at the ECU as possible culprit but it may not even be the problem........we shall see. More parts being replaced meanwhile- cheap/easy. What the heck.

 

[noquacks]

Thanks. As expected, there are a ton of soldered pins on the back of that board. With the magnification limitations I can accomplish on my end, there isn't enough clarity and definition to adequately examine the board for what I am looking for, so it looks like you will have to do it.

It may help if you clean the bottom of that board with alcohol before you start. Start from one end of the board and check each solder joint for failure points around each pin like in the pics I supplied. Note that the pics I supplied for examples were magnified quite a bit, so that should let you know how difficult it may be to find them without magnification on your end. If there are failed solder joints, they probably won't be found on the smaller soldered pins. Like I said, the components attached to heat sinks and power supply-in pins are the usual suspects. They will typically be the larger soldered joints. If you find any, try and focus the camera on that particular area and take more and clear pics before you attempt repair. If you can locate them and it fixes the issue, there is value to others by coming back here and posting your results of the location(s) of your bad solder joint and your success, if any, in the repair.

Additionally, if you want to try again with smaller quadrant pictures of the board and post them here, that may help us as well. But just so you know, that looks just like the type of GM/Delco boards I have experienced the very same issues with before. I've seen it on windshield wiper motor boards. Cruise control boards. Anti-Lock Brake Control Boards, stereo boards and ot

Cold solder joints are rare on ECUs - my experience is that they are much more reliable than most manufactured boards as they are designed to work in harsh environments. A quick 'scan' of the image provided shows no particular issues that I can see and trying to check EVERY soldered joint is asking a bit much.

The single edge connector pad at the bottom left of the picture is VERY suspect though and needs attention - it corresponds the other 'dirty' connections on the other side of that board PLUS the contacts in the socket the board plugs in to (per @Bluejets post above).

Fix those edge connections and you'll probably find everything else works.

Do you have a schematic that might indicate what those connections are for?

I goofed bt rubbing to hard with fine sandpaper on that one edge connector pad which appears copperish color. How stupid of me. I was warned about this from you guys (I read it too late). How bad could this be? Will it oxidize quickly once installed in operation?

 

[kellys_eye]

I goofed bt rubbing to hard with fine sandpaper on that one edge connector pad which appears copperish color.

The other contacts are only 'tinned' anyway so you could use a soldering iron, flux cored solder and 'whet' the exposed copper pad to recover it with solder. If you don't feel confident with doing this - and there's a risk of the copper track lifting if you apply too much heat - then see if you can find someone with soldering experience. Lots of people do have it!

 

[noquacks]

The other contacts are only 'tinned' anyway so you could use a soldering iron, flux cored solder and 'whet' the exposed copper pad to recover it with solder. If you don't feel confident with doing this - and there's a risk of the copper track lifting if you apply too much heat - then see if you can find someone with soldering experience. Lots of people do have it!

Thanks, kelly. Good to know it can be repaired without too much difficulty. UPDATE: The car runs GREAT! What did I do? I popped in a new throttle positioning sensor, even though I checked voltage and it checked out (the symptoms were suspect of a bad TPS). $9.99 and a new TPS isn out of the equation. Then, I popped in an new O2 sensor but NKT brand, Made in JAPAN, no China junk. Wow, I couldnt believe it. Car runs better than before the 3 week hickup. And yes, this is with the computer that was in there that I rubbed to hard on that contact. I will keep an eye on it.

Anyway, I owe many thanks to all here that supported my problem. I will be back, no doubt in the future. You guys are great!