Subaru Outback Forums banner

Code P2138/Throttle position Sensor

305K views 202 replies 70 participants last post by  outdoormilk 
#1 ·
Unfortunately my 2006 Outback/Legacy LL Bean edition has had this problem for 3 months now, it comes and goes, the check engine light comes on and car goes limp, gas pedal stops working, let it sit and it will start and gas pedal works:rolleyes:; dealership couldn't even figure it out...a code that has shown up is P2138 I have heard it is the gas pedal position sensor? Local mechanic thought it was the computer; used one is $400, I am not interested in that option. Any help would be appreciated.:confused:
 
#2 ·
Gotta love drive-by-wire. :rolleyes:

The service manual has an 18-step diagnostic procedure for this code. It mostly comes down to repairing the harness/connectors or replacing the accelerator pedal sensor. Replacing the ECM is a last resort if everything, and I mean everything, else checks out.

:7:
-Jeff
 
#172 ·
Anybody have any more info?

I am currently having the p2138 problem exactly as described. The local dealer says he has never heard of it. I have an 05 Outback XT. Has anyone had this successfully fixed?
Had the accelerator pedal assembly replaced at 20,000 miles. Less then 20,000 miles later same problem again 2014 outback third assembly coming up. Is this a recall item?
 
#4 ·
Same problem, same year, anyone have an update?

I have the exact same problem on the exact same year - 2005 subaru outback XT

The other day it happened when using cruise control which I thought was an interesting data point...

Tom
 
#5 ·
In case anyone else checks this thread - I appear to have significantly reduced the occurance of this problem by seating the connect on the accelerator pedal assembly. I was looking for potential places were there could be a loose connection or frayed wire. I noticed that the plug had backed off a ways on the connector, so I pushed it back down.

Was happening 2-3x per week before doing this. Hasn't happened since doing this, and its been about 4 weeks.

I'm not saying I cured it for sure, only time will tell since it is intermittent.

Tom
 
#35 ·
I just started getting this P2138 issue happening in my 2006 Subaru Leg./Outback wagon last month and as a result am afraid to drive it to work. It happened 3 times in a row driving up or down the 17 mile mtn. road to my job. I'm afraid one time I will get rear-ended on a blind curve. It starts by the cruise-control light blinking (I rarely use cc.) and then engine light coming on. Then after a few minutes the accelorator goes limp on me and I have to restart the car- to reset the computer I guess. Then it works tho warning lights remain on. After a long while and another restart the warning lights go out. Oddly the dealer at first tried to tell me that they couldn't fix the problem because they have been dealing with the exact same issue for the last 2 months for another guy and haven't been able to fix it. He is up to spending over $400 and is now looking at replacing a $1000 part and who knows! A person from another forum said he fixed this problem with an $8 can of lubricant. Any new news on this problem? I have reported it to both Subaru and to the NHTSA as well. I have an appointment with the dealer who wants to keep my car for 3 days just to gather data on it. I have a mechanic (non-dealer) who will try to help me but don't want dealer to tell me they won't help if he can't do anything. All earlier posts seem long ago. Has anyone finally fixed their problem or kept it from coming back?
 
#6 ·
Thanks for posting your follow-up and, hopefully a solution.

The 2005+ fly-by-wire accelerator pedal actually has two sensors that track the position of the pedal, and two circuits going to the engine control module. The duplication is for safety -- both signals have to be within a fixed range (voltage) of each other. If they don't match, the ECM puts the engine into fail-safe mode.

The loose connector is a very strong possibility. It's also possible that in other cases, because the accelerator assembly is at floor level and could be exposed to water, salty solutions and dirt, the contacts themselves could get dirty or corroded, leading to bad connections which would also trigger the code.

Let us know if it does work out.
 
#7 ·
I agree that it could have also been the pedal assembly given it's location and the fact that it doesn't look particularly sealed. I was going to replace that next if this connector reseating didn't work out. I'll post back if it reoccurs, otherwise it looks fixed for now, thank goodness.

Tom
 
#8 ·
Dead Accelerator

I too had the same exact problem. The dealer claims no knowledge of the problem too-what a shocker. They ran my 2006 Outback for 3 days trying to get the problem to reoccur. It never did, but they finally replaced the APP Sensor/Pedal Assembly when they realized that I found it is a prevalent problem-from this website.
It cost me $100 deductible (as I had already purchased a very expensive extended warranty). I am going to try to get this back from Subaru myself. The dealer said he could not get them to reimburse me and they "could not" do it themselves. This is a safety-critical issue that needs to be addressed and Subaru needs to come-clean on it! I will do what I can to get them to do it. I hope it does hot happen to me again while I am doing 65 in the middle lane of a 6-lane highway. Did we really need drive-by-wire technology? Not happy about it in AZ.:28:
 
#9 ·
Please contact NHTSA

I too experienced this problem on 06/19/2011. My 2006 Outback has a service appointment on the 23rd of this month. Please report your problem with the NHTSA at 888-327-4236 or online at safercar.gov. They compile statistics regarding dangerous defects and provide car owners with some level of protection when there are enough specific safety-related complaints. There are now only two complaints on file regarding accelerators that become unresponsive. My family was fortunate to have survived the circumstances surrounding our loss of throttle control and resulting deceleration. From my perspective this is a serious safety issue that Subaru really must address.
 
#10 ·
Fixed Mine

Hi Everyone,

I've been dealing with this problem on my 2004 Outback for the past year. Cleaning and reseating the connector would fix the issue for a week or two, but the problem would always come back. I finally got sick of trying to fix the connector, so I cut it off and replaced it with some quick disconnect connectors from Radio Shack. (3 bags of 1/4" part #6403058). These new connectors have more "bite" into the metal tabs on the sending unit, so I'm hoping the issue won't come back. So far so good.

After I removed the original connector and took a closer look at it, I can see mine clearly has a manufacturing defect. There's an orange gasket around the edge of the connector, but on one side of the plug it is not aligned correctly. I believe this let condensation in and created a bad connection. Condensation makes sense based on the problems I saw, since they would usually happen first thing in the morning or anytime I put the AC on the floor setting.

For anyone else with this issue, you may be able to fix it my simply putting some grease around the edge of the connector and trying seal this gasket better.

Fixed
http://i596.photobucket.com/albums/tt46/cpyne/IMG_2168.jpg

Bad gasket
http://i596.photobucket.com/albums/tt46/cpyne/IMG_2180.jpg


!!!!!!!!!!! UPDATE !!!!!!!!!!!!! - 1 week later and the problem was back. Apparently the radio shack connectors don't make good enough contact either.

Now I've soldered leads directly on to the sending unit connectors.

http://i596.photobucket.com/albums/tt46/cpyne/IMG_2182.jpg
 
#37 ·
Hi Everyone,

!!!!!!!!!!! UPDATE !!!!!!!!!!!!! - 1 week later and the problem was back. Apparently the radio shack connectors don't make good enough contact either.

Now I've soldered leads directly on to the sending unit connectors.

http://i596.photobucket.com/albums/tt46/cpyne/IMG_2182.jpg

Did you have any issues with this after the soldering?

I have been dealing with this in my 06 Outback as well.

Thanks.
 
#12 ·
In regard to another thread, I found that there was a change in the design of the accelerator pedal assembly around MY2006.

Prior to the change, the accelerator pedal used variable resistors (potentiometers) that varied with the position of the pedal. For 2006+, instead of variable resistors, Hall effect IC's are used.

Variable resistors depend on metal contacts that slide on a length of resistance material. The moving contact can wear, and become contaminated/corroded. This could lead to incorrect signals, and the trouble code. The Hall effect system doesn't depend on moving electrical contacts -- instead it uses magnetic fields, which should be far more reliable.

opposedforces.com lists two p/n depending on date of production, for all 2.5 Legacy (Outback) from 05 through 08. 36010AG021 appears to be the earlier version, and 36010AG140 is for later production.

I wonder if the replacements being supplied for defective pre-change pedals now use the newer Hall versions . . . .
 
#17 ·
I was having the p2138 issue on my 2006 outback. I cleaned the throttle postition sensor connections on both the throttle body(under the hood) and the one on the pedal with alcohol. After cleaning and reseating, I had no issues for 6 months, then just recently the p2138 check engine reoccured. I am debating replacing the sensor on the pedal or the whole pedal assmebly. The dealer quoted just the TPS part at $189.
 
#19 ·
Subaru tech support has told us of some 05, 06 and up to 09's having a issue with the wire harness under the intake. With the scantool on the default angle with koeo should be 4-5 degrees; moving the harness can cause a fluxuation in the readings.
 
#20 ·
Interesting. Did they say what the problem is in the harness? Sounds like harness wires breaking or shorting.

I would imagine (hope) this will lead to issuance of a TSB. Although, from the experiences reported above, the DTC P2138 seems to be resolved by replacing the accelerator pedal assembly. Perhaps the harness problem leads to a different DTC.
 
#21 ·
Mine resolved after reseating the connector on the pedal. Well worth the 1 min it takes to try it.

Tom
 
#22 ·
Cant rule anything out; we just had a issue with the code, and had a pedal in stock. The Senior Master Subaru Tech had me google this code as we hadn't seen it before. So I found this link and we discussed it. He called Subaru Tech Support and they told him of this issue; so with the scantool in place he wiggled the harness and saw the numbers change. Harness on order; will let you know how it turns out.
 
#23 ·
. . .we just had a issue with the code, and had a pedal in stock. The Senior Master Subaru Tech had me google this code as we hadn't seen it before. So I found this link and we discussed it. He called Subaru Tech Support and they told him of this issue; . . .
Is the problem with the car you have in now consistent, or intermittent? If it is constant, i.e., always triggers the P2138 after the code is cleared, then the wiring harness replacement might be conclusive. But if the DTC isn't consistent, then it might take some time to prove this out.

. . .so with the scantool in place he wiggled the harness and saw the numbers change.
Which could well indicate a problem in the engine wiring harness (assuming that's the harness being wiggled/replaced) which connects to the throttle position sensor.

However, I was looking at the wiring diagrams and there's nothing to suggest that the connections between the accelerator pedal to the ECM go through the engine compartment wiring harness. It's all in the cabin bulkhead harness. (Whereas, the throttle position sensor wiring, of course, starts out with the engine wiring harness and connects to the bulkhead harness going to the ECM.) So it will be especially interesting if replacing the engine harness resolves a P2138. It would mean there's a lot more to this code than appears to be in the FSM.

Thanks for raising this possibility.
 
#25 ·
I'm thinking they are wiggling the connect down by the peddle. I think there is an issue down there with the exposure to the elements (water from people getting in the car with slush on the boots) that cause problems with the contacts, but that's just my suspicion.

Tom
 
#26 ·
That would make sense given the experiences reported here; but I'm looking at "under the manifold" and the KOEO default angle. This seems to refer to the throttle (on the engine intake) which, under these conditions, does default to a roughly 5% position. The accelerator pedal has a zero deflection scanner reading at KOEO when the pedal isn't being pressed.

But I'm hoping it will be clarified; i.e., which harness, and which reading on the scanner.
 
#27 ·
Any vehicle using a drive by wire system will throw a "correlation" code. It means there is an issue between what the pedal position sensor is sending and what the TPS inside the t-body assy is sending. The various codes are ambiguous in all manufacturer testing. The only vehicle that has been consistent in a particular part failure with similar correlation codes has been the Dodge 4 cylinder model throttle body assembly. The motor takes a dump and won't move. On Fords, the throttle plate sets at the sprung position of 20 degrees and won't move. This is because when the motor fails it can't hold it in any position so the built in spring load holds it.

All other brands have been various issues from loose connection, bad ground, to failed sensor or t-body motor assembly. I have only had two Subaru problems and the cars were within warranty, so I sent them to the dealer. Both customers came back and reported. One a loose plug at the pedal. The other was a dirty throttle body. Both had the same code.
 
Top