Electrical 1995 GT - CEL/MIL shorted to SIGRTN inside the ECU itself

taz95gt

New Member
Dec 5, 2018
3
0
1
Croatia
Hello everyone,

I'm having a set of very weird symptoms that I didn't really find a lot of info on the internet, I'm trying to see if there's any electric and/or electronic experts here that could help me out.

I think my ECU is fried, but I don't know how or why, and I want to make sure it's safe to replace it before I do.

Symptom #1:
Upon turning the key to the on position, the KOEO codes begin flashing, and upon starting the car, the KOER test begins. The problem is, the STI connector is not jumped to the SIGRTN port on the adjacent data link connector. I can't turn it off.

Symptom #2:
MIL/CEL randomly flashes during driving, before the period where it would turn on to continuously glow, as it usually should. However, I didn't see any rpm drop or hear any noise that could be attributed to a misfire when it happens. This also occurs for the O/D light, but is not synchronized with the MIL/CEL light.

Symptom #3:
I'm getting a code 543 for the fuel pump control circuit not supplying power to the pump. Thing is, the car can start, run and drive just fine and there are no codes 556 or 557. It does occasionally drop rpm (once, some time after start) before returning to normal for the rest of the drive (be it 15min or 2hrs long), which could be attributed to fuel pump randomly losing contact, but it hasn't stalled once.

Symptom #4:
I'm getting codes 172, 173 and 176, though those could simply be old and slow O2 sensors.

The only other code is 332, but that's probably due to the TAB and TAD being deleted by the previous owner.

As for the first symptom, I have traced it to a series of what I think to be shorts within the ECU. I've took the ECU out of the car and performed these measurements on my desk, with it entirely disconnected.

I've measured the resistance between the following pins:
46 (SIGRTN) and 48 (STI): 4k ohms​
46 (SIGRTN) and 17 (MIL/CEL): 4k ohms​
46 (SIGRTN) and 26 (VREF): 670 ohms​
8 (Fuel pump power check pin) and 26 (VREF): 6k ohms​
8 (Fuel pump power check pin) and 46 (SIGRTN): 5k ohms​
Many other pins near and around 26 (VREF) have around 22k ohms between each other​

This leads me to believe:
A) The 4k ohms is enough to start KOER and KOEO tests​
B) MIL/CEL is shorted to at least one pin it's not supposed to short to (in this case STI)​
C) SIGRTN is shorted to VREF​
D) Pins at and around VREF are either corroded or the isolation between them is burnt through to cause these connections.​
E) Symptoms 1, 2 and 3 are all caused by these shorts​

Furthermore, there is no visible or discernible damage on the connector, it's pins, the board, or the places where the pins are connected to the board, which brings me to my question.

Question:
What could be the possible cause of this damage? Could high voltage from a defective sensor or harness burn through the isolation on the board?

Could corrosion happen on the inside of the board and connect these pins, even though there is no visible corrosion on the outside, or on the leads going into the board from the back of the connector? The board is translucent, and when shining a light through the part where the leads meet the board (on the inside side of the connector) I see no deformities or impurities in the plastic that look like corrosion.

What other causes could be out there? Could this just be regular wear and tear?

I want to make sure before I put in a new ECU that it's not going to get damaged as well.
 
  • Sponsors (?)


You've done far more analysis than I would have. :) I don't think there's any data out there on what the various resistances between pins is supposed to be, so I'm not sure if that's really very useful information to have. It's unlikely that your ECU is 'fried' - the fact that it's running, driving, and returning various codes accurately tells me that it's largely functional. These 90's ECUs were actually remarkably durable, and serious problems with them are rare.

I'd attempt disconnecting the test-pin wires from the ECU harness to eliminate shorts / grounds within the wiring before suspecting the ECU itself has a problem. Far more likely that those are corroded or shorted. Also check all your engine and firewall grounds. Nothing causes more oddball ECU behaviors than a missing or corroded ground connection.

Momentary (flashing CEL) warnings are normal, and are often related to fueling issues (possibly O2 sensors or the car not being in a good state of tune). An occasional stumble or stall can also happen due to the same root cause, as our cars are prone to hunting & surging at idle.

The 543 could be a fuel pump going bad, or a problem in your fuel pump wiring or harness connections. That'll get worse if that's the case. A lot of these issues can be caused by faulty O2 sensors or a problem in that wiring as well (as it runs near the exhaust over time a short / melted harness is common).

With that all said, it absolutely could be the PCM. But unfortunately that's usually the last thing to troubleshoot. With the exception of a poorly mounted add-on chip, obvious corrosion, or perhaps a bloated or blown capacitor from visual inspection.