I bet no one has seen this!

[KFXGUY]

So bare with me because I?m going to include every detail I can think of.
2004 ls1 gto. Mods are really unimportant.
So intermittently I?ll have an issue where the tps (cable drive) % won?t return back to zero. It?ll start out getting stuck at .4%. Then for a little bit it?ll get stuck even higher at .8%. Sometimes it?ll work itself out. But other times it?ll start this cycle and work it?s way up to 9%. I?ll sometimes plug my scanner in to watch this. And when it gets stuck at 9%, I can press the throttle and it lags way behind, like there?s a huge delay and the engine will barely run or might not even run at all. The only way I can fix this is to disconnect the tps and iac, turn key on and let it sit for about ten minutes. Then plug back in and start. Wtf. It did this to me once while dropping my kid off at school in the drive through line. Quite embarrassing.

So here?s what I?ve chevked/changed
Grounds. All appear to be good. I even added another one.
New battery
New alternator
I?ve changed the tps about 5 times. New delco units.
I had a China throtttle body. I changed it to a nick Williams.
I did relocate the ecm inside the vehicle. It did this before however.
Randomly I?ll find random tables that zeros themselves. Wut? Like yesterday, my boost map was zeroed out.
I did have headers on the car that burnt through the heat shield protecting the ecm a while back. I don?t recall if this happened before that. The computer has been flashed many many times.

This is the most odd problem I?ve come across and I don?t find anything on the web like it. Anyone have this happen?

[KFXGUY]

Something I want to add when this happens, the tps voltage does not go up with the percentage. It stays the same as it’s supposed to.

[kingtal0n]

some guesses to get you started:
A. bad ecu
B. poor TPS wiring
C. poor ecu grounds
ex.
voltage can be induced (inducted? "cross" emf or "cross talk" one of them fancy words) from one wire to another when they are very close or if the length is very long (this is why some manufacturers use 'shielded wire'). I saw a car once that when the gauges were turned on via a switch, the tps and map would all go crazy.
The TPS gets its ground and 5v power from the computer, so my guess would be the source (perhaps the ECU plug is loose) or the length (of wire) is getting interference from another wire, or is simply poorly grounded. Or bad.

The idea of it going up as temperature increases is fairly standard for electronics with resistance to ground issues. I.e. the hotter the ground wire gets, the higher the resistance, the more voltage 'stays' in the sensor wire, the higher the TPS% will read.

[HaCKoWeLL]

Do you have a Fluke meter or access to an accurate DMM? I would check for voltage differences across the ground points before you go any further, especially after relocating and wires being burned through. You cant just check for voltage to a specific wire/device because until its under a potentially higher current draw it may not show. I deal with this on a daily basis working on vehciles that someone else has "Added" stuff to before it came to my shop, but I do see OEM issues also..especially on older wiring or stuff that is exposed to the elements/moisture. After checking grounds then I would check how much voltage you are losing to the 12v Batt and Ign feeds, at the ecm. If all that checks out then move to the 5v sensor supply wires, but remember all this needs to be checked under a load to find the problem.

What normally happens is someone checks for 12v+ with a cheap DMM or test light and thinks they are fine and so they go chasing ghosts..but under a load it will be a different story. Glass AGU fuses are especially horrible because the solder on the endcaps will pass enough current to keep the voltage up near normal until you hit a few amps of draw and all the sudden you see 7-8v on the meter and the fuses looks normal not blown. Just like any other bad connection it will cause intermittent issues that drive you crazy, but it looks completely like a normal working fuse.

If you put your DMM on DC volts and put the red lead on one end of a wire and the black lead on the other end at the battery/power source, you will get a voltage reading hopefully near 0v. As you increase the load/amp draw on this wire, the voltage will increase..and this is how much voltage you are dropping/losing. Things like too small of wire, length of wire come into play with large circuits/high amp draw stuff like amplifiers but for an OEM wire in your vehicle it shouldnt ever be over .5-1v under load. The higher the number the worse. You can also check the difference between two ground points to see if you have a heavy enough ground/bad ground the same way.

If you end up with a wire that comes suspect, inspect it from end to end and at both terminations. If you dont see anything it is probally broken inside even though it looks fine, you can find this by probing through the insulation. This happens on quite a few DBW harnesses near the connector where it is stressed, and also a fair share of MAF harnesses. Good luck and let me know what you find.

[KFXGUY]

some guesses to get you started:
A. bad ecu
B. poor TPS wiring
C. poor ecu grounds
ex.
voltage can be induced (inducted? "cross" emf or "cross talk" one of them fancy words) from one wire to another when they are very close or if the length is very long (this is why some manufacturers use 'shielded wire'). I saw a car once that when the gauges were turned on via a switch, the tps and map would all go crazy.
The TPS gets its ground and 5v power from the computer, so my guess would be the source (perhaps the ECU plug is loose) or the length (of wire) is getting interference from another wire, or is simply poorly grounded. Or bad.

The idea of it going up as temperature increases is fairly standard for electronics with resistance to ground issues. I.e. the hotter the ground wire gets, the higher the resistance, the more voltage 'stays' in the sensor wire, the higher the TPS% will read.

Thx. Is does get worse after driving a bit. Well sometimes.

[KFXGUY]

Do you have a Fluke meter or access to an accurate DMM? I would check for voltage differences across the ground points before you go any further, especially after relocating and wires being burned through. You cant just check for voltage to a specific wire/device because until its under a potentially higher current draw it may not show. I deal with this on a daily basis working on vehciles that someone else has "Added" stuff to before it came to my shop, but I do see OEM issues also..especially on older wiring or stuff that is exposed to the elements/moisture. After checking grounds then I would check how much voltage you are losing to the 12v Batt and Ign feeds, at the ecm. If all that checks out then move to the 5v sensor supply wires, but remember all this needs to be checked under a load to find the problem.

What normally happens is someone checks for 12v+ with a cheap DMM or test light and thinks they are fine and so they go chasing ghosts..but under a load it will be a different story. Glass AGU fuses are especially horrible because the solder on the endcaps will pass enough current to keep the voltage up near normal until you hit a few amps of draw and all the sudden you see 7-8v on the meter and the fuses looks normal not blown. Just like any other bad connection it will cause intermittent issues that drive you crazy, but it looks completely like a normal working fuse.

If you put your DMM on DC volts and put the red lead on one end of a wire and the black lead on the other end at the battery/power source, you will get a voltage reading hopefully near 0v. As you increase the load/amp draw on this wire, the voltage will increase..and this is how much voltage you are dropping/losing. Things like too small of wire, length of wire come into play with large circuits/high amp draw stuff like amplifiers but for an OEM wire in your vehicle it shouldnt ever be over .5-1v under load. The higher the number the worse. You can also check the difference between two ground points to see if you have a heavy enough ground/bad ground the same way.

If you end up with a wire that comes suspect, inspect it from end to end and at both terminations. If you dont see anything it is probally broken inside even though it looks fine, you can find this by probing through the insulation. This happens on quite a few DBW harnesses near the connector where it is stressed, and also a fair share of MAF harnesses. Good luck and let me know what you find.

Good info. I’ll dig a little deeper.
Btw, the wires or nothing was burnt. The header tube rubbed/burned a hole through the ecm heat shield so I suspect it got the ecm much hotter than normal.

[anniversaryss]

What kind of throttle body?

[KFXGUY]

What kind of throttle body?

I had a China 92. Bought a new nick Williams. Did the same thing. I think it’s the ecm. The bad thing is I gotta get a new ecm and Pay for two credits again.

[matty b]

Sounds like the PCM to me.

[anniversaryss]

I had a China 92. Bought a new nick Williams. Did the same thing. I think it’s the ecm. The bad thing is I gotta get a new ecm and Pay for two credits again.

Sigh.... one thing that sucks about this

[KFXGUY]

Sigh.... one thing that sucks about this

Yea I’m not sure why they can’t transfer my license over to the replacement ecm. Not fair really on my end of it.

[anniversaryss]

This happens with a lot of swap people, especially if VIN swaps are needed iirc.

[KFXGUY]

This happens with a lot of swap people, especially if VIN swaps are needed iirc.

Well I haven’t swapped anything yet. This is the original ecm in the car. I never mentioned that so I figured I would.


Couple things I noticed. I can do a tps reset. Problem goes away until I flash the computer several times. I’ve been making a lot of changes lately and that seems to trigger it. It was running perfect except my afr was way too lean in boost. So I made some dueling chainges only past 100kpa (which has nothing to do with idle) and this time when I flashed it the tps got stuck and then unresponsive and starting surging and died. I’m not so sure if my computer has any relation to it. Seems like when this happens, the scanner will lock up and I have to start the program over. I’m getting really agitated.