dimswits wrote: ↑Mon Mar 29, 2021 11:59 am
Your intuition about the ECU ground is on the right track, but likely the wrong end. If you read through the thread Brad linked where I posted my findings, Bob Weir has a couple of posts later on that get to what I believe is the root of the matter... the wiring harnesses in both cars were never very good, and age is only exacerbating the matter.
Going back to my Turbo Scorpio, I'll add this seemingly unrelated anecdote. I have a voltmeter in that car, and it would routinely read less than 10 volts when coming to a stop with the lights on. The lights would dim and then brighten again. It would often, though not always, read only 11.9 to 12.1 when running. Sounds like a bad alt, huh? But I never had issues starting the car.
So I wired up two more voltmeters, one directly to the battery, one to the alt. For good measure I put a plugin voltmeter in the cigarette lighter. And I went for a drive at night and did some observing. When my cluster voltmeter dipped below 10, I looked at the batt and alt meters. Both read 13.6. The cigarette lighter meter read 11.9. WTF!?
After a bunch of probing around the car and getting various readings between awful, low, and good, I narrowed it down to the three main splices in the body wiring harness, S1002, S1003, and S1004. Since 1004 is the only one remotely easy to get to under the dash (the other two are under the fuse box), I opened the harness there.
I should have taken pictures, but the splice was gross. The factory stripped a long length of the voltage source wire, wrapped it around a bundle of 3 branch wires, held it in place with a small crimped ring, then wrapped the thing with electrical tape. The exposed wires had significant white and green powder corrosion, despite being inside the car.
On a whim I clipped the two red wires that branch away from the splice. One feeds the ignition switch, and from there a whole bunch of stuff. The other feeds the headlamp switch. I ran each through a fuse link direct to my power distribution point.
The result was no more flickering lights, and no more readings of less than 13V on my voltmeter when running.
I think this, along with the observations from my datalogging, and Bob's findings when exploring the harness, show that the wiring is prone to deterioration in the power distribution splices that isn't obvious, and that can cause varied and unpredictable voltage drops. I think that simply redoing the voltage supply to the eec with new wiring of sufficient size, using the old relays to switch new ones fed directly from the battery, will mitigate the high idle issue.
