Yes, it's a much discussed topic. Want to see if I am on the right diagnostic track.

Car (a '94) began acting odd at start-up. It would crank and once it fired there would be an odd grinding sound as though the starter was not disengaging properly. I observed this a few times so not just a one-off. Fair enough, I assumed the starter was on the way out (never been changed to my knowledge) so it was time.

Very shortly after the engine began having a hunting (loping) idle at start-up (between 1000 / 2000 after it warmed up a bit). Idle became very erratic in general. On most cold-starts it would fire up and be at around 600 (operating temp idle speed). If you shut the car off, it might not start at all ... foot on the gas - it would start and run. Other times it would start up and go higher (1100) and just stay there, not dropping as the engine warmed up. Occasionally it would just stall out.
Diagnosis - ISCV or ECU.

Pulled out ECU. Found One cap had leaked out (I've seen worse) but some electrolyte got to the board. Other caps tested good. Replaced physically damaged cap, left others in place (they tested good - but i still plan to replace. Did not have the others "in stock").

Car started right up, seemed to run better on test drive, even had more power.
Next day, back to the same symptoms. Erratic idle, stalling out.
Next step - ISCV.

Removed, cleaned, rebuilt (new bearings) valve. Coil windings tested good, operation good with direct battery voltage.
Re-install valve, engine starts right up. Went through a few start / stop cycles to re-set valve.
Same symptoms again (loping idle, no start without foot on gas etc).

BUT, now i notice something different. On shut-down, valve makes loud buzzing sound instead of the usual clicks. Not every time, but sometimes. I heard this twice.
Makes me believe the issue is in the computer. That valve was very carefully gone through and tested (by me). ECU controls the valve so ...

Also noticing a fuel smell (like it's running rich) which also makes me think ECU or maybe it's just dumping fuel trying to keep the car running when it's about to stall out?

Anyway,

Thoughts?

 

There may be not a single problem going on. At least the starter motor's pinion gear is not retracting well enough. It should retract fully as soon as the key is turned from the start position to the on/run position. Aged starter tends to behave that way due to the badly clogged sliding gear shaft.

The second problem may be related to the leaked liquid remaining inside of the layer or surface of the ECU board or traces may be partially damaged. The liquid becomes solid and conductive and makes the current paths. It's not easy to remove it especially the inside one. I too sometimes find the incorrect mounting of the capacitor's +/- reversed.

If I were you, I'd rebuild the starter and also check / clean the ECU board.

 

Caps are installed correctly. ECU has never been opened until now, only the physically damaged one has been replaced and the person doing the work is of serious pedigree.
How many layers are there in addition to the top and bottom? It was mentioned to me by my electronics "guy" that this could be an issue. He wanted me to try it as is before digging deeper into it.

 

The board has 3 layers and the center layer is mostly used as a grand plane. It has been 30 years since the board was manufactured and other than QAS capacitors, there are aged stressed parts. Metal resistors tend to increase the resistance. The electromigration / ionic contamination may occur. It would be the best to try a good working one if possible.

 

Yeah, this is kind of where I'm at. Plugging in a known good one would be the definitive diagnosis (if it ran right of course)
Your input is very much appreciated!

 

OK.
Thanks to a very generous local SC owner that let me borrow his computer to try in my car ...
The results are in.
The car started, warmed itself up and idled down flawlessly on the first turn of the key.

Fires up cold at around 1100 rpm, idle gradually drops at coolant temp rises,
once it hits operating temp idle falls to a rock-solid 700 rpm and stays there.

Tried multiple hot starts (often a big problem) and once again ... flawless behaviour.
Quick flare up over 1000 and quickly falls back down to 700 and sits there.

Long story short, I need an ECU or to get the one I have fixed.
I'm guessing that one cap that leaked out has damaged something (maybe on the middle layer) on the board.

Has anyone out there successfully made or had a repair like this done?

 

Thanks for the update. You were lucky that there was a very generous SC owner near you. I believe that you were found to be a good citizen by him too.

To purchase a good working ECU is the quickest way although it's not very easy to obtain as well as it may be expensive.

If I were you, I'd start checking the surface traces again very carefully using a good magnifier and an ohm meter aiming to check the continuity completely. The ECU board is coated by two different materials. One is the green coating which is aimed to protect copper traces from the oxidization/corrosion and the second one is the conformal coating to protect parts and the board from the moisture. The leaked QAS often expands under the green coating and it eats up copper traces. Sometimes parts around the QAS capacitor are damaged. The QAS eats up the solder too and you have to be careful to check soldering points around the QAS capacitor.

In my past experience to fix electronics units requested by dealer managers and shop owners, almost 95% of boards were fixed. I needed to give up those rest of 5%. Those are caused by un-obtainable semiconductors and badly damaged printed circuit boards. Some boards were badly burnt and some internal layers were damaged.