2004 TL 6MT. I'm just starting to diagnose and issue that popped up yesterday. I was about a 1/2 mile into my commute to work when the VSA icon, VSA message and CEL all came on and car went into limp mode. I puiled over, restarted and it drive fine the rest of the way to work w/ CEL on.

Pulled codes when I got work:

PGM FI: P2138 - APP sensor A-B correlation
VSA:

• 25-1 - yaw rate sensor failure
• 61-1 - battery voltage failure
• 83-1 - ecm/pcm relation failure

Battery voltage = 12.67V power off, 14.2V running, so I'm ruling that out. Codes cleared. Drove about 1000 feet and back into limp mode. Re-pulled codes and only P2138 and 83-1 returned. Every time I shut the car off, VSA light and CEL are on, but throttle works normally until I drive a short distance. I'm unable to turn VSA off using the "VSA off" button on the dash.

I was doing some research here last night and found a thread that mentioned moisture dripping onto the ECU from condensation on the AC ducts above it. I did encounter an AC issue on a trip 2 weeks ago. My wife and I had a dramatically different temps set on the dual climate mode. About 2 hours into the trip, It was getting hot in the cabin, air was barely blowing out of the ducts and I the aix mix doors would not change between defrost/foot/head, etc. When I parked the car at our destination about an hour later and left it, a GIANT pool of water was left under the car shortly after. Ice dam?

Anyway, I exposed the ECU this morning. Bone dry in and around the area under the dash, BUT, I'm pretty sure that the dust on top of the ECU looks like it water had dripped on it and dried. See photo below.

I'm going to try and limp this thing home in a few minutes so I can work on there over the weekend. My plan is remove the ECU and take the cover off to see if it's possibly still damp inside. Just hoping that, if this is ECU related, it isn't damaged.

 

If it's the ECU the one below is the cheapest one on ebay:
2004 -2008 Acura TL manual transmission ECM PCM ECU computer 3.2L J32A3 MT J32A | eBay

 

Make sure the ECU you buy is 2004-2005. The clips changed in years 2006-2008, and wouldnt be compatible with yours.


looks like water drippings on the ecu. Did you first try to remove it and see if there is any water in there? the top cap isnt water-resistant.

 

The P2138 is the app sensor. You can change it yourself.

that should fix the limp mode issue.

 

Appreciate input thus far. Sorry I haven't posted back yet. I got the car home and put in the garage and have only put in a couple of evenings of diagnostics since.

- I pulled the ECU, removed the cover and let it sit overnight in case there was any moisture inside. There were no indications of water intrusion, but it cost me nothing to try it. Reinstalled, no change. I'm assuming ECU is OK.

- Further testing pointed pretty clearly to APP sensor, so I ordered a new one. Installed it and there was ZERO throttle response. I then pulled P2127 - APP sensor B voltage low. When I stream the live data from APP sensor, sensor A is showing 0.902V, sensor B is showing 1.725V.* Checking for continuity across terminals 1 and 3 on the new sensor, it was confirmed DOA.

- I put the ORIGINAL sensor back on the car. Throttle now works correctly with no DTCs now for PGMFI systems. Traction control light remains on all the time. VSA light comes on after I begin driving, but does not go into limp mode anymore. The VSA DTC has now changed to 84-1 - Sensor logic failure.

- I attempted the VSA Neutral Memorization reset, but cannot got it to complete. I pulled out the VSA Off switch, tested it and found no continuity between terminals 3 and 4 on the switch, so it appears the switch is bad. I was going to try and swap it for the on in my Pilot, but checking the Pilot service manual, they are not wired the same.

Is it possible that a bad VSA Off switch could be causing all of this?



*one other note above APP sensor A/B voltage - the service manual clearly indicates these values should be the same, that's why I replaced the sensor. I streamed the live data off my sister's 2005 TL though and they are also not the same, but that car has no issues or DTCs. The A voltage on that car is higher than the B voltage. The A voltage on my original sensor was higher than the B voltage. On the new sensor that didn't work, A voltage was lower than B voltage, confirming a problem with the resistor in the B side. Having the service manual contradict what I'm finding on a working vehicle isn't making this any easier.