99 Altima hesitation: I think I'm onto something!

In my *limited* experience, the semi-short is probably AT THE CONNECTOR ITSELF, where the wires join up with the connector pins.

You will notice that unless the connector is potted, and I bet $2 of your money that it isn't, some adjacent wires are shorting via the connector itself.

TAKE APART THE PLUG, and inspect it with a magnifying glass, especially the wires/solder-terminals around the wires you're concerned with.

The solution is to pot the wire/solder connections, or just bend them out of eachother's way.

Lg

By "potted" do you mean that each wire enters the connector in its own channel and doesn't share a common space with the other wires? If so then that's actually how the connector at the TPS is made. Also, the two wires are not adjacent at the TPS connector, they have a third wire between them, and there is no conductivity between the middle wire and the outer two.

The two wires ARE adjacent at the ECU connection however, but no amount of prodding or jiggling seems to produce any variation in the resistance... I would expect at least some change if they were making partial contact there. Then again, I would also expect some variation at some point as I've been wiggling the whole harness from the TPS to the ECU, but I haven't had anything definite.

I haven't been able to get either wire out of the ECU connector yet, it looks like I have to insert something in the terminal side (not the harness side) of the connector in the hole above where the pin from the ECU goes, and pry the locking tab down, but that hasn't worked, yet.

At this point, I'm seriously considering cutting one of the two wires at both ends, and jumping it with a new piece from the TPS to the ECU... the only issue there being that that becomes a weak spot in the connection and I would have to find a way to run a wire permanently.

Thanks for the reply, I'll keep at it...

It doesn't have to become a weak spot, it can become the strongest part of the harness. When I do such things myself, I take a piece of air hose of the right length, slit it on one side for the length, and

*tuck* the new wires into the rubber hose, then tape the hose closed along the incision that I made. A utility knife with a razor blade works best for this. Just make sure the hose/wire assy doesn't come into physical contact with hot parts, like parts of the exhaust system, and route -away- from electrically noisy things, like plug wires and coil packs.

Silicone rubber in a tube can seal up the plug itself and make it waterproof.

Lg

Good point. It's not a very high resistance short in relation to the resistance expected at WOT though... 1.2 Kohms vs 4, so it seems like a short of that level would have an effect.

I'll give that a try and see what I get.

Thanks for those suggestions, I will test all of those things and post back.

Hi Kevin,

I just read your other post RE: Testing automotive electronics, and realized you may have thought I was testing the TPS directly and seeing the short between two pins on the TPS itself. What I'm actually seeing is a short between two separate wires in the harness between the ECU and the TPS, with BOTH ENDS of said harness unplugged. So in a normal situation, there should be NO conductivity between the two wires.

I also measured the TPS resistance at the ECU with the TPS end of the harness connected, and the ECU connector disconnected, and this was where I was seeing about 1/3 of the expected resistance at WOT.

Well, actually it is between the signal ground wire and the reference voltage, but yes. In my original post I was thinking it was between the TPS signal & the signal ground, but that's not the case.

I've got the factory service manual and it doesn't show any other devices connected in parallel, at least not in the TPS section of the book. Maybe one of the overall schematics would tell a different story. If there are multiple devices sharing the same reference voltage signal I would expect they are tied together inside the ECU and not all joined somewhere inside the harness?

I'm 99% sure I have the right wires; I can test either at the TPS end or the ECU end and I get the exact same resistance between the supposedly unattached wires, and if I test from one end of either wire to the other end of the same wire, I get a hard short, so that would seem to indicate I have the right wires. I did think it was strange that it's not a hard short between the two wires though, and also strange that it does not vary at all, even with all the moving I did of the wiring harness. The most change I saw was maybe .03 Kohm and I couldn't ever be sure that was due to my fiddling with the wires or just random changes in connection strength.

I'm anxious to give that a try in the morning and hopefully learn for sure if there is a problem here or not.

Not at all, it just sounded as if you had misunderstood my original post so I wanted to clear that up.

And I appreciate that. As does the rest of the group I'm sure.

Anyway, I'll do some testing of the voltages in the morning and report back; thanks again for the input.

Now that I think about this, something's not adding up. Maybe I've got it backwards in my head after going over it so many times. But if the semi-short is between reference and ground, then theoretically the current should pass through to the TPS rather than cross the 1.2Kohm bridge, unless the resistance between the ref. voltage and signal ground is greater than

1.2K. Obviously a test of voltages will show the truth...

If I might jump in with a suggestion here. Measuring resistance can be a bit tricky due to the ability of moisture, oil, and even dust to carry a reading across your high impedance volt meter test points. These ghost shorts may be able to carry the low current from your DVOM, but would not be able to carry enough current to affect you TPS signal adversely. It would be better to measure the actual voltage signal at the PCM harness. If you back probe Pin 23 (TPS signal) at the PCM harness connector with the connector plugged into the PCM and the ignition switch on, engine off, you should read 150 mv to 850 mv at closed throttle. As you depress the accelerator, the voltage should climb steadily to a wide open throttle reading of 3.5 v to 4.7 v. Having an oscilloscope would be nice because you could watch the voltage trace for any glitches, but if you are very observant and steady you may be able to catch any glitch by seeing the voltage fail to climb steadily on your DVOM. If your meter has selectable voltage ranges, use the lowest ranger possible to view 5 volts. If it has an analog display be sure to watch it carefully.

If you do not get the expected voltage ranges then be sure to verify that you have a good 5 volt reference voltage at the TPS and then perform a voltage drop test on the TPS negative wire to ground. Any voltage reading across the ground return indicates unwanted resistance that needs to be corrected.

Just testing resistance readings is not always a very conclusive diagnosis for TPS circutry.

Lets see if I am getting this right.. With both the ECM and the TPS disconnected from the harness. You are getting a reading of 1200 ohms between the TPS signal wire and the Referance voltage wire? You should get an infinite reading during that test, unless (and I would have to check a wiring schematic) there is some other circuit in parrallel with the TPS on your vehicle. Keep in mind that the Reference voltage wire is connected parrell to many other sensors. Only the Signal wire (to pin 23) is a direct wire from the TPS center pin back to the ECM. Do you have a good wiring diagram? Look it over carefully and be sure you are testing the correct wires.

Well, I can see where that might be a problem, but don't be fooled by some continuity across the circuit due to possible contamination. That sort of short will not usually carry enough current to skew the TPS signal measurably, but it will show up on an ohm meter. If there is indeed a hard short across the two wires the reading would usually be closer to 0 Ohms. There again, be sure you are testing the correct wires. Double check. Still, checking the signal voltage is a much more reliable test with less chance of error.

My post on testing wasn't aimed at you, I just thought it might be some helpful information to other readers. I hope you didn't take it as a slight on you capabilities. The truth is, you seem to have a pretty go grasp on the situation. Anything I can do to help is offered gacefully.

There in lies the problem. The reference voltage (sensor power supply 5 Volts) wire leaves the ECM and branches out to all the sensors that need reference voltage. The signal ground also leaves the ECM and branches out to all the sensors that have a signal ground. You are reading the parrellel resistance of all the other sensors in the system if you check from reference to signal ground. You would have to disconnect every sensor to show an open circuit between those two wires. You need to be checking from reference (ECM pin 49, Red wire) to TPS signal (ECM pin 23 , Yellow wire) to find the specified 4,000 ohms at WOT. It will all make sense to you if you could look at the whole wiring diagram and get an understanding of the circuitry.

Kevin

BTW, Engine performance electronics is my favorite type of work.

Maybe this will help.

Here is a link to a good explaination of a TPS It is Generic in nature but does apply to your type of TPS.

Kevin