Electrical project

[bwana50]

Confession up front: my level of electrical knowledge is roughly "connect the hot wire and the ground wire, and the light turns on" - hence I turn to the omniscient forum for electrical engineering help.

My wife's Durango has been having issues that seem to be the PCM - we're on #4 now. I've read these are a problem, but have also read that other issues may simulate a bad PCM. It's a bastard to diagnose because sometimes, after it dies, it will immediately re-fire; sometimes it might be a couple hours. It might be a two months until the next episode, or two hours. I haven't yet found any indicators that might be a problem outside the PCM. Never any error codes stored.

Here's where the engineering part comes in. I'd like to build a "diagnostic indicator" that I can splice into the harness to help rule out other issues. What I'm envisioning is a small panel with three lights that I can rig in the "cockpit":
(1) one that shows if the 5V sensor power loop has been shorted;
(2) one that shows if 12V power to the PCM is lost;
(3) one that shows if ground to the PCM has been interrupted.

I think that, for #1, since a short would be power-to-ground, that I would run a hot wire to the light, and the "ground" wire to the light would tap into the 5V positive line. Yes? Would I need a diode on that?

For #2 and #3, it would be fairly simple to rig it so the light is 'on' normally, and goes 'off' if the condition occurs, but I'm trying to do the opposite - light goes on when the connection in question is broken. How would I do that? Now that I think about it, it would probably be OK to wire it the 'easy' way - but see the next condition.

Finally, how could I rig the system so that the light stays on (or off as the case may be), even if the problem goes away? For example, if it's an intermittent grounding problem, if ground is lost the light comes on but stays on even if ground is restored?

[crash]

I'm not an EE either, but unless you can look at loads, as in the amount of current draw, and build something with thresholds, you aren't going to get what you want here.

You must remember that every complete circuit is not only a positive connection but also a ground connection, so the proposed solution will not do you much good unless you can determine when the flow of electrons is "out of spec" and the circuit is shorting.

[Chief]

Crash kind of nailed it. If it's just an intermitant short or open that is only there long enough to set the condtion that faults you car, it may not be enough to trigger a latching circuit to keep the "fault light" lit.

A latching circuit/relay is what you are looking for to keep the light on after the fault goes away. But as I said, there are a lot of conditions.

Case in point. I had a cracked coil pack/spark plug on one cylinder. It would misfire intermitnantly. You could feel it when sitting in the car. But the diag machine could not capture the fault as the amplititued/duration of the misfire was not enough to trigger a fault OR for it to show up on a diag machine, and it would not show up as a misfire fault on the ECM. It was finaly found by watching a scope and seeing the misfire on that cylinder.

[crash]

Originally Posted by Chief

Crash kind of nailed it. If it's just an intermitant short or open that is only there long enough to set the condtion that faults you car, it may not be enough to trigger a latching circuit to keep the "fault light" lit.

A latching circuit/relay is what you are looking for to keep the light on after the fault goes away. But as I said, there are a lot of conditions.

Case in point. I had a cracked coil pack/spark plug on one cylinder. It would misfire intermitnantly. You could feel it when sitting in the car. But the diag machine could not capture the fault as the amplititued/duration of the misfire was not enough to trigger a fault OR for it to show up on a diag machine, and it would not show up as a misfire fault on the ECM. It was finaly found by watching a scope and seeing the misfire on that cylinder.

Excellent point. We run some pretty fancy computers on our race stuff, but these types of conditions are still very hard to diagnose. We have thresholds for faults so that a wild whatever that occurs once every 40,000 revolutions doesn't neccessarily throw a code or put it into limp mode, but the factory ECMs, while I am sure have these parameters, are usually much more lenient on allowed fault conditions before throwing codes. This means, as you said, that even if you have the technlogy to detect it, if it is set in a manner that it desn't find your particular problem, then it is as good as worthless.

Electrical problems are a real PITA. Sometimes you just have to keep throwing new parts at it until you find the problem part. Very frustrating.

Good Luck!

[Cobra Cory]

Yeah I know that some folks have had good luck taking a marine bilge blower and putting a hose on it from the grill to the PCM box like the cool air setup in my Roadster, or moving the PCM away from the heat of the engine.
It seems the heat is what is killing the PCMs. You would need to wire it in so that it starts blowing as soon as the truck starts.

I also know that changing the CPS, (Crank position sensor) under the right fender liner helps as well. Sometimes it is not the PCM but the CPS which acts the same as a bad PCM.

Another trick is when it shuts off to wiggle the rear plug going into the PCM if it starts it is the plug wiggling out. USe some zip ties to pull the plug into the PCM tighter. Also put a small piece of aluminum foil in each of the plugs holes. This makes a better/tighter connection for the pins. just don't short them out on each other. Some times you just have to try a little Yankee ingenuity.