P2138 and Reduced Engine Power

[Cbrowne]

Hello all,

Camaro owner from the UK experiencing a fair few issues right now. Firstly, the car has previously had a P2138 which was resolved with a new APP (accelerator pedal position sensor). Since then the issue has come back around with the exact same P2138 and reduced engine power, despite my disbelief of my almost brand new APP going bad I got a new one anyway, which did not resolve the issue. The P2138 and reduced engine power will only come up once I hit the kick down part of the gas pedal, although, 1/10 times the fault will not flash up, the other 9/10 times are the problem. Since then I have now replaced my throttle body and even wired in a new throttle body pigtail (I am an electrician). I know everything has been installed correctly but I am still encountering the exact same issues. It is also worth noting the issue will come up whether the engine is on or off. Can someone provide any advice on what else to check as I don't fancy taking the chance of spending £600+ on a new PCM/ECM. If anybody has links or PDFs to wiring diagrams from the PCM to the throttle body and pedal that would be grand as Google doesn't seem to be very helpful right now. Any help would be much appreciated.

Thanks all!

[gtstorey]

Full diagnostics are available for a reasonable fee from alldata.com or with an acdelco tds subscription (assuming these are available outside the US).

I think the P2138 is one of the tricky ones that is checking correlation between rising and falling voltages across both the accelerator pedal sensor and the throttle position sensor. And if its an intermittent problem, it will probably be be hard to track down the problem. And next to impossible to track down without something to pull freeze frame data from when the code triggered.

Do you have a throttle commander or similar connected or has the car been "tuned"?

[Cbrowne]

Originally Posted by gtstorey

Full diagnostics are available for a reasonable fee from alldata.com or with an acdelco tds subscription (assuming these are available outside the US).

I think the P2138 is one of the tricky ones that is checking correlation between rising and falling voltages across both the accelerator pedal sensor and the throttle position sensor. And if its an intermittent problem, it will probably be be hard to track down the problem. And next to impossible to track down without something to pull freeze frame data from when the code triggered.

Do you have a throttle commander or similar connected or has the car been "tuned"?

To my knowledge the car is completely stock and has had no flash or tuning done nor throttle mods. The fault appeared after regular driving and no matter the weather has stuck around for the past 2 months whilst waiting for parts. I'm thinking it could possibly need to be plugged in and the throttle relearned on a scan tool but I've read they're usually pretty good for relearning themselves. I have completed the bog standard relearn procedure of 3 mins on, 1 min off and repeat to no avail. There are no obvious signs of corrosion on any connectors but without knowing the pinouts I couldn't even begin to test cables other than looking for the 5V reference voltage which is present on both APP and TB. I will look at calling around for some auto electricians to have a look as they must have better kit than me, but I usually get turned away as they refuse to work on imports.

[gtstorey]

How long have you had the car? A lot of driving and then out of the blue problem or is this a recent purchase that has pretty much done it from day 1.

[Cbrowne]

Originally Posted by gtstorey

How long have you had the car? A lot of driving and then out of the blue problem or is this a recent purchase that has pretty much done it from day 1.

I've owned it for about 14 months now, fault code appeared about 2-3 months ago out of the blue, done track days and road trips but nothing particular comes to mind occurring within about a month of the fault coming up.

[gtstorey]

This will probably take using the factory diagnostic steps from the sites I mentioned to determine what the problem is. If the fault is constant, it shouldn't be to hard for someone with a decent diagnostics scanner, wiring diagrams and factory diagnostic steps. Without those, probably impossible. But you might be able to do it with a DVM, depending on the exact diagnostic procedure.

[Mouse330]

Is there anything connected to the APP circuits? like a pedal commander etc....If there is disconnect it and re-test, There's a 40 cavity connector X104 located under the brake booster area, Its rectangular, Pull that apart and see if there is any water intrusion, corrosion etc.... The other that's also a 40 cavity rectangular X102 is located RF, I think its beside the PDC(underhood fuse block). Also if u have a DVOM u can check the circuits starting at the APP sensor connector. You would check for 5V reference on C(TAN) and F(WH/BK) KOEO. Check for less than 5 ohms on A(Purple) and D(BN). For reference A is purple and F is white/black, Its a 6 cavity connector. Likely have high resistance in a circuit which would affect a voltage/low reference circuit.

[PaulsSS]

1st thing I would do is remove all the connecters at the PCM and look for corrosion and reseat everything. Next would be to reseat ALL connections in that circuit and look for corrosion-damage. Then maybe move on to testing wiring for damage. The PCM could also be the cause but I wouldn't replace it until you test everything 1st.

[Mouse330]

Any word on a fix or progress?

[Cbrowne]

No fix as of yet, my local garage tested my pedals and my pinouts and found the pedals to be faulty, to which I tested them myself and they all tested no issues at all. I've even tested supply and feedback voltages at the pedal with a digital voltmeter which measured no issues. I ordered another off-brand GM pedal which once again I've tested as working but still have P2138. I have found by resetting my ECM (disconnecting battery) the car acts right for up to about 30-45 minutes before going back into fault when hard accelerating. Graphing the data D/E voltages on the pedals it shows they're working absolutely fine so I can only guess the formula to calculate P2138 has corrupted in some kind of way. I'm now killing the car over night to see if it can completely reset the ECM before I go to the local tuner and see if they can program P2138 out of the car. And for reference 4 OBDII scanners have all proven the only fault is a P2138 D/E voltage correlation.

[Cbrowne]

A bit of an update, I have now found after killing the car it operates with no issues at all for all of about 30 minutes. No amount of driving with a lead foot will bring up the P2138...until 30 minutes later. After the 30 minutes the issue comes back. What I have found is; in the 30 minutes my D/E voltages correlate exactly as they should. I.e:

Sensor 1 20% = Sensor 2 10%
Sensor 1 40% = Sensor 2 20%
Sensor 1 60% = Sensor 2 30% etc.

However, after the 30 minutes, it goes to:

Sensor 1 20% = Sensor 2 12%
Sensor 1 40% = Sensor 2 24%
Sensor 1 60% = Sensor 2 36%

Although, once this has set in, it stays like this until the ECM is reset. I've tried to test the connectors/looms by wriggling them vigorously to see any voltage drops/increases on the D/E graphs but they show nothing out of the ordinary. So far it seems that once the ECM is set in it cannot read the voltages right, I have proven that there is a constant 5V to each sensor with a DVM and proven a consistent feedback voltage is outputted through the APP sensor. Although, without a proper voltage % table I couldn't tell you if it is correct or not, all I know is it's consistent.

Right now I'm guessing there can be 2 causes, a voltage is being induced into one of the cables from the loom, or the ECM itself is creating some kind of interference once fully energised.

What does everyone else think? I'm going to test the circuit resistance once again to see if anything has changed, but from what I remember from doing it a while ago all was fine and bang on specification.

[Cbrowne]

I've done some more investigative work and have a few measured values.

Firstly, each cable from the APP to the ECM is reading 0.2 Ohms, no shorts and no open circuits.

Secondly, the outputs from the APP are as follows:

At 0% throttle - Sensor 1 = 0.958V and Sensor 2 = 0.482V 99.9% accurate
At 100% throttle - Sensor 1 = 3.852V and Sensor 2 = 1.924V 99.9% accurate

However, the ECM is reading:

At 0% throttle - Sensor 1 = 21% and Sensor 2 = 10% 95% accurate
At 100% throttle - Sensor 1 = 79.5% and Sensor 2 = 42.75% 93% accurate

What I believe is happening is as sensor 1 is increasing if it doesn't quite reach the 90% accurate threshold of sensor 2 then it will go into P2138 fault. Next step will be to power up a programmable 5V power supply and look for any voltage drop between the pedal and the ECM plugs.

What I have also found by shorting out the APP is that sensor 1 can only reach a max of 90% whereas sensor 2 goes straight to 100%. In addition to this, sensor 1 doesn't go straight to 90%, it typically reaches 85% before incrementally going to 90%.

It is definitely leaning towards my ECM going bad so will get some quotes through for a clone and looking at swapping them out.