really weird tps issue... (RESOLVED)

[EfiniX]

I've just had a new 4g63 dropped into my vr-4. Once back together the car tuned with a dsmlink v3 without issue. However about 10 miles onto break-in, the idle started acting up. Issue presented as an IAC problem. The IAC value on startup would hit a random number between 88 and 101 and freeze there until the car was turned off. Cold/warm/driving, whatever value initially displayed would remain. Adjusting the BISS showed no change.
I pulled the ECU and inspected it for damage. Caps were replaced 85,000 miles ago. ECU looked fine, with no chip bulges, leaking capacitors, or scorch marks. This was to the naked eye, so grain of salt, but it looked (and smelled) ok.

I ordered a new IAC and installed it this afternoon. IAC values now fluctuate, but generally remain at 101 while driving and during idle. When engine breaking and very occasionally, at idle when warm, the value will fluctuate and appear normal, though never doing below 27.

What is new is the TPS output. TPS while closed shows a throttle position of 84%. The TPS voltage shows 4.28 while closed. The TPS was adjusted properly during the initial tune and the position marks made against the sensor and throttle body still line up (so it hasn't moved).

So, first question, does this sound like something you have seen before? Second, if the TPS had gone south and is showing grossly high voltage while the throttle is closed, would that throw the IAC value of as well? I'm trying to determine if there is a correlation, or if these are separate issues.

My mechanic (EHMotorsports) is going to be diagnosing the TPS tomorrow, but I thought I would check with the community. I've already got a message out to the ECM tuning guys.

 

[GSTwithPSI]

I'd grab my laptop and open up ECMlink. Enable the TPS adjustments (check box), then mess with the TPS offset and TPS scale just to see if you can bring it back into spec. If you are still getting erratic readings no matter how you adjust the values, than I'd bet your sensor is sh*t. It's an easy way to do little testing without even opening the hood.

Bad wiring, or a loose connection could also cause issues. Those are easy checks as well.

 

[EfiniX]

I did try to make TPS adjustments, but DSMLINK only lets you tweak the idle (closed) position if you are within a few % points of 0. I'm so far out so spec, the software spits back an error. Makes sense since the TPS should be calibrated based on voltage with physical adjustments under the hood.

Or perhaps I'm thinking of something else?

We'll see what Evan finds, but you are on the money with our thoughts.... Either some hacky wiring harness BS, or the sensor decided to finaly give up the ghost.

 

[GSTwithPSI]

I think we are talking about different things...

Turn the car to run, but don't start it. Then, start an ECMlink capture. Put the pedal to all the way to the floor, and then let it return to rest. This will log a full TPS sweep. After you have a sweep recorded within the log, ECMlink should be able to auto adjust the TPS. To do that, go in and check the box that says "Enable TPS Adjustments". After you do that, return to the log and see what your TPS is reading. It should be near zero at rest, and near 100% at WOT. If it is not, adjust the scale and offset until the TPS sweep reads 100% at WOT, and 0% at rest. If you can't dial it in, then you have a problem. Typically, not being able to dial it in would indicate it is too far out of manual adjustment. But, since you are confident it isn't, not being able to dial it in or an erratic reading would indicate a bad TPS. If your readings are way off, but are consistent, your TPS might be OK. Even though the sensor hasn't moved, it may just need to be adjusted manually again.

You TPS being all out of wack is definitely causing some idle issues as well.

 

[EfiniX]

Well here's the thing... All this got done (the adjustments) to spec when the engine went in. Something has changed in the last week. I did replace the IAC, but that's pretty straight-forward. The ECU looked (and smelled good). Haven't heard back from the ECMTuning guys, but that's forthcoming. Ultimately, I'm wondering (and in many ways hoping) that the stock TPS w/ 177k+ worth of use has just kicked the bucket. If so, I'll buy a new one and use the downtime to ask EH to see if he can chase down my radiator fan always being on.

 

[EfiniX]

And speak of the devil... Just got a reply from ECM tuning!

Keep in mind that the logged ISC value is just want the ECU is asking the ISC to do. The ECU has no idea what the ISC is actually doing. It's not "reading" an ISC position. It's just showing what the ECU is trying to tell the ISC to do.

A high TPS reading might be causing any number of other issues and it might be caused by something else that could cause any number of other issues as well. I generally recommend that people start fixing the things they know are wrong (like the TPS reading in this case) and often the process of identifying what's wrong there will help with everything else.

Thomas Dorris
ECMTuning, Inc.

...and my response

You have 100% echoed the comments from Evan (and myself). He’s going to start with testing the TPS at the TB, then work his way back through the harness to the ECU. That’s great info to know about the ISC. I always though that the ISC value was what was being read, but if that’s what the ECU is telling the ISC to do, then it’s far more reasonable to think that a faulty TPS voltage would/could result in the ECU telling the ISC to do something at idle that it should only be doing at high throttle.

Thanks!

C

 

[r4pt0x]

Have you measuered the output Voltage when the TPS value is at 84% (or another weird value)?

I once had to deal with an Eclipse D20 with multiple failures - one of them was the TPS sensor that went stuck at random positions after a few minutes of driving. After hitting the gas pedal a few times it loosened and went back to normal operation for a few minutes.
The ISC was also dead - this can be easily indicated without any further equipment by removing it from the throttle body and turning on the ignition. It should move ~1cm out and back in. If it doesn't try another one.
If the next one also won't move, grab a scope and check the PWM signals at the ISC and at the ECU pins.

I also once had an ECU with one blown transistor for the stepper motor driver. It's the same symptom as a "half dead" ISC: The ISC rattles a bit (as one coil is still working), but it won't move in/out. This was quickly discovered by checking the PWM signals at the ECU pins and then switching to another ECU...


BTW:
The ECU increases the ISC steps on WOT. So if the TPS gets stuck at 84% it would be OK for the ISC to be set at a high step position too! So I would check the TPS first before suspecting the ISC...

 

[EfiniX]

Well Even got deeper into checking the wiring and such after the TPS checks all passed. Then he got to thinking "hey, resistance changes with heat".

Sure enough when the car started warming up, the TPS lost its mind. Ordering a new one today!

 

[coyotes]

Evan is a pimp, no question about it. /ubbthreads/images/graemlins/cool.gif I would have lit the car on fire before I figured that out.

 

[EfiniX]

Well as it turns out, lighting the car on fire would have helped to diagnose the issue.

 

[EfiniX]

And with the new TPS....


I'm in business!

DSMLink = first mod ANYONE should do. No replacement for being able to collect info from the ECU.

And seriously if anyone lives within 500 miles of EHMotorsports, he's the absolute greatest!

 

[coyotes]

Good to hear!

and yes I plan on giving evan all of my dollars. It's too bad he's way down in portland, he needs to move up to seattle or buy a car trailer. /ubbthreads/images/graemlins/grin.gif

 

[EfiniX]

Don't you dare! I love English Racing, but I may have been able to save my #3 if they had told me they had to tune my injectors @ 850cc (they are 950cc) to prevent the car from running too rich. I'm guessing they figured I didn't know what I had and really, without pulling the injectors out, how would you know? Top notch guys (and having spent a good 30 minutes talking to Lucas this morning while trying not to drool over a 1350hp GT-R he had just finished for a guy in Texas), I got a better sense of his process. However, if you really want to get under the hood with the mechanic and learn a bit along the way, you need a guy like Evan.

He's just terrific.