I had an 'epic win gearhead moment' June 29, 2016 - I'm reposting it here.
"Gearhead moment.. So, I went to look at Curt and Ronee's 2002 Escape ( V6 auto AWD ) a couple of days ago. Code was P0352 coil on plug cyl #2. We changed the coil last year and Curt changed the injector on #2 last month - didn't fix the problem. Concerned that a broken wiring harness or PCM were the next steps. I pulled the wiring diagram. Figured out how to test the primary sides of the coils in situ from the PCM to the noise supression capacitor. No broken wires - or odd coil primary readings.. So.. It's in the EEC-V PCM.
I proceed to read a little bit about this problem.. Coil drivers in the PCM go bad throughout a few years production of Ford Escape... I look around the wrecker knowing full well that i can't use a computer without the dealership programming station to reflash (reprogram) the VIN and anti-theft keys. Open up a couple of similar (but not identical) computers. Look at the circuit. Guess that the six driver chips were for the High Current Ignition stuff becuase they're in a modified TO-220 case and there are 8 pads on this EEC-V where they can be soldered to - and only 6 are present.
I proceed to take the PCM apart at the wrecker, pry the circuit board off the heat-sink, break it cleanly at the first row of pins at the 104 pin connector and use a pair of tin snips to harvest the driver chips from a couple of the boards. 12 driver chips in all.
My dad's friend from work Mr. Jay Henselwood (sp?) (20-30 years ago) told him that you take the chips off a circuit board with a blowtorch, if you're just saving the chip. So i did that at my folks place and had 12 loose high current drivers. I made the mistake of attempting this in their basement. I made this mistake for only one chip. It turns out that all the (mosfet) driver chips had the same part number. A Proprietary part number that I can't get a data-sheet for.
I put an ohm meter on the patient's PCM and could not find continuity from any one driver chip to the PCM pin. I looked at the wiring diagram and made a couple of guesses - based on the following facts:
- the three wires that drive the bank with the bad cylinder all go in along one edge of the PCM. (the converse with the good bank and the other side of the PCM is also true) so, It's probably one of the drivers on the side where the wires go.
- Cylinder 2 between 1 and 3 in the bank.. so I guess the driver chip is probably the one in the middle. I reach in with a small cutter and snip the legs off it and start to de-solder it and it cleans up nicely.
I try to de-solder the heat sink. NO WAY.. I try the propane torch.. the conformal coating is starting to char and things are getting sideways fast. the board is heat sinked to the cast pot-metal chassis with thermal bridging silicone. I can't get it hot enough to solder and i've already cut the presumed defective critter off at it's knees. I use a screwdriver to GENTLY lift the circuit board off the chassis/heat sink wile Curt softens the 'glue' with a propane torch. first one side - then the other. It comes free of the chassis. I clean off the adhesive with a razor knife gently trying not to nick the foils. the blade touches a suspect foil that feels like it lifted when it was heated and i'm thinking - it's over. it's a $300 touch at the stealership... oh well - pessimistically thinking 'might as well close it up and see the new error codes'.
I pick one or two driver chips out of the plastic bag. Check them, then check two or three more with a meter - open circuit.. all directions.. What?! ok.. gamble.. They can't all be bad - I'm going to solder this little guy (chosen arbitrarily) on the board and I solder it down. The solder joint is homely but bright and contiguous.
I explain that I have to replace the thermal silicone and I remember that I had a tube that I put in a box 15 years ago with some other electronics supplies. I rummage at my parent's place and locate the tube of glue. The next day I re-glue the computer to the chassis and reassemble the PCM. The Instructions say that the cure time for the glue is 24 hours. My buddy Evan is eye-balling my homely solder joint, and not saying anything but clearly communicates his dissatisfaction with my technique with a sideways glare. He's fun that way. I proceed to concern myself with the hypothetically ruined PCM and how hard it will be to buy knock off programming tools from China and break into the software on a junkyard PCM, reprogram anti-theft keys etc.
Curt phones me shortly after 5PM the next day - two hours ahead of schedule. He says, " I reinstalled the computer and it works. No Check-engine light, No stumble, spun the tires, went up the hill like Jack the Bear. I had goosebumps. Don't let anyone tell you that you can't solder." Ronee also drove it and said it's better now.
I honestly didn't think the repair would work. I made more guesses and used more picnic table techniques than I would normally be comfortable with. I took a swing at a problem that I thought I could handle, then decided halfway through was impossible and thought it was not going to work at all. Well, in this case at least, I guess not quitting before completion was the real key.
Who replaces a defective COP (coil on plug) driver IN an EEC-V to avoid reprogramming it... (with a blowtorch... and two soldering irons.. on a picnic table FFS) ? Me, That's who... I'm incredibly pleased with myself and the fact that I kept another robot-car out of the dealership."
F.T.W. :D