NOT Normal P Lock Malfunction - ONLY Happens Over 80 Degrees. Never When Cooler

If PPOS is not just point A to point C then all of those other connectors must be released before the "check for short" will be a valid test. Because other paths to ground may exist normally.

I think that none of the other connectors must be released for the "check for open" path. Because even though there are other connections, none of those result in an alternate current path through the ohmmeter. The path between A and C may be carried on more than one wire in the cables, but as long as there exists at least one complete path between them the test will pass and the cable should work.

If you still get different resistances to ground from A and C then put the positive lead on ground and the negative on the pin and measure again. Because that is strange and changing the polarity may tell us something.

Also measure resistance A to C. If the short to ground accompanies a break between A and C then different resistances to ground from those pins is possible. Imagine a circuit which is A-B1-R1-B2-C and B1-R2-GND and B2-R3-GND. (This is a triangle B1-R1-B2-R3-GND-R2-B2 with paths A-B1, B2-C coming off two of the vertices.) If R1 is zero then you should be measuring the resistance to ground of R2 and R3 in parallel from A or C, but if R1 is large or infinite (open) the A-GND and C-GND can have different values corresponding to R2 and R3, respectively for infinite (open) R1.

 

Throwing ECUs at it isn't the way I'd proceed, but it can be hard to give ideas about diagnostic technique over the internet.

You've already posted freeze frames showing collapse of the supply voltage reaching the transmission control ECU at certain moments. If it were me, I'd be pursuing that lead aggressively.

In an older vehicle, I once wired an LED into a suspected intermittent circuit and taped it to the dash, so I could drive around and just see when the LED went dark.

The same thing might be useful here. A question could be: is the voltage reaching the transmission control ECU really dropping, or is there some fault inside that ECU and it's imagining that the voltage drops? A light independently wired in there would give you an independent answer.

There are electrical testers that make a buzzy noise instead of a light, so your eyes don't have to be on them. They can be hooked up to some wiring while you concentrate on shaking it and stuff.

The first page of a DTC section in the manual will give a list of several different things as possible trouble areas. That doesn't mean to just assume any of those things might be causing the problem. The list will include crazy unlikely, but just conceivably possible, causes: the kind you don't start fussing with unless there's evidence you've picked up while troubleshooting that clearly points there.

 

OL on meters means "out of range". That is, the value is too large to measure. In this case that means the resistance to ground is huge, as it should be if the insulation is intact. Be aware though that you would also see OL if the wire near the pin being tested was broken with a gap introduced, or even if the ohmmeter lead wasn't plugged in.

Historical trivia. There is some disagreement if OL stands for "Open Loop", "Over Load", or "Over Limit". OL arose when displays used 9 segment LED numeric digits and the only symbols that could be made were composed of combinations of those segments. The maximum value in the significant digits was 1999 (for a four digit display). If the meter needed to show 2100, for instance, it could not, so all such values were shown as OL. Meters with more sophisticated displays today will often show a phrase like "over range" instead.

 

Or use a cheap DMM (Harbor Freight sells one for $7, sometimes less) set in voltmeter mode and tapped into the questionable 12V wire. That way you can also see if the voltage starts to decrease slowly as it approaches the failure point, or if it just collapses all at once. If you can get access to at least a cm of bare wire perhaps use a posi-tap - it only makes a pin hole in the insulation and can be removed (and reused).

Posi-Tap- No Crimp Tap

If there are multiple wires you want to monitor, you could use a multiposition rotary switch, run one wire to each position, and the meter to the common. Something like this:

Rotary Switch - 10 Position - COM-13253 - SparkFun Electronics

Just be sure that whatever switch you use does not short neighboring pins as it rotates. It shouldn't, but check before using it.

 

Rotary switches are made both ways, shorting on purpose, or not shorting. So, the kind that shouldn't, shouldn't.

 

I had to order positaps from Amazon or Ebay, don't recall which. Possibly fakes, but they work. None of the local car parts stores had any. Note that they come in different sizes, pick the one designed for the gauge of wire you will tap. (They cover a pretty wide range.)

I would start with whichever wire supplies 12V to the TCU. If the wiring harness is of the type where the wires near the plug can be moved apart some then you can tap the wires there. If that isn't possible you might be able to back probe the pin at the connector. This guy makes his own back probes:

or you can pick up a kit, like this one:

Back Probe Kit, 22-Piece

I have never driven around with a back probe in place, but if it is secured tightly with several zip ties it shouldn't fall out. It might however lose connection to the back of the pin. Be very careful that the back probe itself cannot short to anything. Tapping the wire would be more reliable.

 

No, it is no conductance, the exact opposite.

Think about it - if there is no path for current to flow between the two probes then the measured resistance is infinite. (Air isn't quite a perfect insulator, but for the purposes of this discussion let's say that it is.) The meter can't tell the difference between "infinite resistance" and "resistance too large to measure", and in all such cases it shows "0L".

There is only a short in a cable if the connectors are all unplugged and the measured resistance between a pin and ground is a small number or zero. That is "0", not "0L". This concerns a short from the wire directly to some piece of metal attached to the car's body. There are also shorts that go from one wire to another wire and in that case the low resistance path that results would be from one pin to another pin, usually on the same connector, but not always so in wiring harnesses that fork.

 

That sounds promising. At this point you need to figure out exactly where that ECU is so that you can aim directly at it. The ECU will have a circuit board in it, and broken solder connections can be temperature sensitive. The solder cracks but maintains an electrical connection until thermal expansion opens a gap in the crack. If you can make that component fail with applied heat, but it (temporarily) recovers if you tap it (flick it with a finger, if it is too far away for that, tap on it with a dowel or some other nonconducting rod), that would be a very strong indicator that the ECU is bad, and that the problem is a bad solder joint. Tapping it may or may not allow recovery of the hot ECU, but blowing ambient temperature air from the dryer should. In short, if you can drive that ECU in and out of failure with different temperature air applied only to it, then that ECU is almost certainly the problem.

 

correct

nope.

When checking for an open circuit between pins A and C the resistance between A and C are measured. Because the resistance from one point on a wire to another point on the same wire is typically very low, the spec calls for a value of less than 1 Ohm. Some specialized ohmmeters may give an underflow error on a value which is too close to zero, but the odds are that isn't what you are using. They usually just round 1 microOhm to 0.000 (or however many significant digits are available). 0L would indicate high resistance in this scenario too. If you suspect that somehow your ohmmeter gives 0L on a very low resistance circuit do a positive control - just touch the two leads directly to each other. You know that the resistance in that circuit is tiny so the result should be zero. If the meter keeps giving 0L when it shouldn't in resistance mode suspect a broken wire in one or both of the probes. Cheap probes often break like that. Also, there is a battery in the meter, be sure it is still good. Like a Prius, a multimeter can do odd things when the battery goes low. Although, thankfully, most put up some sort of low battery warning.

Summary that might be useful (cable unplugged at all ends, test only at pins A,C and at a ground):

A-C (no shorts, no opens) A or C to ground is very high resistance, A to C is very low resistance
A-B-C (short at B to ground, no opens) A or C to ground is low resistance, A to C is very low resistance
A-B1, B2-C (no shorts, open at B1/B2) A or C to ground is very high resistance, A to C is very high resistance
A-B1, B2-C ( B1 and B2 shorted to ground, open at B1/ B2) A or C to ground is low resistance, A to C resistance is at least
the sum of the A to ground and C to ground resistances

very high resistance will often display as 0L, it might also show as something like 1.999 10^9
very low resistance will usually display as 0.000 (the number of digits varies between devices)
any other value will be some number larger than zero
on auto-ranging meters also have look for a "k" or "M" or some other indicator of the multiplier to see if it is 1,1000, of 1000000 Ohms
on non auto-ranging meters set the appropriate range for the value being measured. (That, is don't try
to measure 50 Ohms in the 200k range, it will usually round down to zero even if it is really 50 Ohms.)

Luckily <1 Ohm is usually good enough for dealing with car wiring problems, and we don't care if it is 0.1 Ohms or 0.001 Ohms. Because accurately measuring very low resistances is very difficult. See for instance:

Accurate Low-Resistance Measurements Start with Identifying Sources of Error | Tektronix