pswayze

A month ago, my backyard mechanic replaced the alternator belt on my 2003 Civic LX 5 speed. It drove for about 30 minutes and then when into extreme limp mode. No tach, no engine temp, running awful so I limped it home. While doing some reading about the issue, I found out that these are symptoms of fried PCM. Improper tightening of the alternator mounting bolts can cause a floating ground for the alternator. This floating ground starts causing the voltage output of the alternator to increase until it feeds back through the PCM. This kills the computer. I found the alternator wiggling around in it's location and when I removed the alternator I found both alternator mounting bolts have stripped out the threads of their mounting holes. One stripped hole in the alternator mounting bracket bolted to the engine. The other stripped hole is in the alternator body itself. Here's what I've done so far:

• Bought and installed a new alternator bracket which was a chore
• Re-tapped the threads in the alternator housing
• Dissolved all of the aluminum stuck in both bolt threads with muriatic acid
• Took the alternator to O'Reilys and they told me it tested good (I'm skeptical)
• Cleaned off all electrical contact points of alternator
• Re-installed original alternator
• Tracked down a compatible PCM and installed it
• had a mobile locksmith flash the immobilizer/PCM
• installed a new battery

The key pairing step was successful but the car immediately threw the P0563 code. The car runs and drives fine. I can clear the code but whenever the car is started again, the code comes back. I followed this procedure from the service manual:

• Turn the ignition switch OFF.
• Disconnect ECM/PCM connector E (31P).
• Measure voltage between ECM/PCM connector terminal E7 and body ground.

I have 12.8V at E7. That leads me to:

• Reconnect ECM/PCM connector E (31P)
• Measure voltage between ECM/PCM connector terminal E7 and body ground.

I still have 12.8V at E7. That leads me to:

• Disconnect ECM/PCM connector A (31P).
• Measure voltage between body ground and ECM/PCM Cconnector terminals A3 and A2 individually.

There is no voltage at A3 or A2. The instructions are to substitute a known-good ECM/PCM and recheck. If symptom/indication goes away, replace the original ECM/PCM. I just bought a replacement. Now I'm being instructed to buy a second replacement just to test. I'd really prefer not to do that. Any ideas?

ezone

Do you know how to backprobe a connector so you can measure voltage while the harnesses are connected and everything is LIVE and the engine is capable of running?

I'd probably want at least 2 voltmeters, maybe 3 just to be able to watch several things at once.

Monitor...... E9, A2, A3.

E9 should become battery voltage while the key is in run and cranking positions.

A2 and A3 should both show identical voltages whenever the main relay #1 is turned on, basically whenever the ignition switch is turned on those should show battery voltage.
Key on = 12v
Key off= 0v

If A2 and A3 stay powered up even though the ignition switch is turned off, that's a problem. Now you need to see why.

If they don't power up when you turn the key on, that's a problem too.



Could be a main relay stuck on, could be wiring issue keeping it turned on, could be internal PCM problem either keeping the relay turned on when it shouldn't be, or misinterpreting power signals for those terminals.


If the wire terminal for A2 or A3 has broken, bent, or lost its tension, the resulting poor connection could be causing the problem.




How to get around programming.....
I'm reading that this problem occurs while the PCM is supposed to be off, and it turns on the CEL after the next key cycle...so it would appear you don't need to be able to actually start the engine to reproduce the code if the fault is present.

Know anyone with an identical car (01-03) you can swap parts from for testing?

Can you swap the PCM into another car and see if you can cause the code to set?
Can you swap their known (assumed) good PCM into your car and see if you can cause the code to set?




Kinda random thoughts here. Hope they make some sense you can use.

pswayze

I can collect a few multimeters. I was backprobing the harnesses anyway as my multimeter probes were too wide for the female side of connector.

I am unclear about your statement. Are you assuming that this problem only occurs when the PCM is supposed to be off? Or that is a fact of this code? What can we deduce if we don't have to start the car to trip the code? Just to clarify, the car starts and runs fine.

Unfortunately I do not have a car I can swap into nor do I have an absolutely known-good PCM. This PCM is surprisingly uncommon with a manual transmission and Canada emissions so even the one replacement I found was difficult to come by. Even if I did, I would hesitate to put it in the car.

I'll build a truth table for E7 E9 A2 and A3 for OFF, RUN, and CRANKING. Is that reasonable?

Thanks a ton for your help!

ezone

I was reading some service information, wondering how taking everything apart to measure is gonna let you figure out what went wrong while it was all together. I figure checking while everything is hooked up and operating might reveal what the error is, if it's there.

Assuming? In order to set the code it looked to me like the IG1 has to be low while the PCM was still awake and operating. What do you think would cause this?

pswayze

I have absolutely no idea. I wasn't second guessing your instructions, I was just trying to better understand what the code meant. The only description I've been able to find is "unexpected voltage" which I thought meant systemvoltage>17Volts. That's why I found the service manual instructions really confusing and unrelated. What you're suggesting is a mismatch between IG1 and PCM states which causes the diagnostics steps to make more sense.

ezone

You can get in trouble by thinking LOL

From what I gathered,
IG1 (E9) was below 2v which means the key had to be turned off,
yet the PCM remained powered up by circuits IGP1 (A3) and/or IGP2 (A2) when it shouldn't have been. This had to last at least 4 seconds.

The PCM can't turn on the CEL in this {partially powered up} state, so it waits until the next 'key ON' to illuminate the CEL.
(this is why I figured it may not actually need to run before you can see the code)




Or something else strange occurred to make the PCM think this happened.

I hate flowcharts and trouble trees with a passion.
I read what info I can find and get out wiring diagrams and try to figure out the 'why' parts long before I resort to actually following them.

pswayze

You are correct that it throws this code even if I turn it to RUN but havent started the car.

It doesn't look like the forum will let me generate a table. These are values with connectors still plugged into PCM.

OFF:****E7(12.6V)*****E9(0V)*****A2(0V)*****A3(0V)
RUN:****E7(0V)*******E9(12.6V)**A2(12.6V)***A3(12. 6V)

ezone

If those voltages all switch pretty much instantaneously when the key is flipped on and off (time is critical here) YET the code still sets and you proved there is no power going to A2 and A3 after shutoff, I'd suspect the new (used) PCM is bad.

Weird failure and the PCMs are pretty reliable (except for the loose alternator issues or direct shorting of output drivers), so I'd sure want to test it in another car and/or vice-versa with a good one in your car before buying another.

Did you get it from a wrecking yard and will they warranty it?


Were the part numbers of the old and new PCMs an exact match?

pswayze

I will have to repeat the test to see if those switch instantaneously.

I did get it from a wrecking yard. I doubt they'll warranty it but I can check.

In desperation, I went to AutoZone and bought a replacement IG1 main relay. I installed it and the light P0563 came on as usual. I started the car, sent "clear all ecu faults" via bluetooth OBDII, and drove it around. I turned the car off and the code did not come back. I did multiple combinations of turning the car off, on, and driving it around, for about 20 minutes. Code has not come back yet. I have no idea what the problem was, or if I've even fixed the problem. The car drives great and even if the PCM is wonky, I'm going to drive it until something bad happens.

ezone

So a new relay fixed it? Pretty sure I mentioned that as a possible cause in the first reply.

Leave it alone and see how long it lasts!

pswayze

You definitely did mention it and as it was the only thing that I could replace for

NDNV

Amazing advice ezone! You rock bro!


I hope this backyard mechanic isn't a relative or friend though....