1999 Legacy Outback DTC P0106 problem HELP ME (very long)

[jamesc]

1999 legacy outback 2.5l giving me fits, owned car a couple years. 6 months ago CEL comes on with P0106. this is the ONLY code car has ever thrown

visual inspection checks OK and no driveability problems

replaced MAP with boneyard part, no change

inspect vacuum hoses no apparent problems but replaced anyhow (checked filter)

checked TPS adjustment per friend (previous Subaru tech)

bench checked sources solenoid, checks OK

peformed solenoid test with underdash connector, checks OK

 

gave up for months now it's due for inspection

CEL light will sometimes go out

checked continuity/for grounds on all three lines from MAP to ECU, all good

checked MAP voltage, 5.1 reference, 3.75 KOEO, 1.49 at idle, specs call for 3.9-4.1 KOEO and 2.0-2.3 at idle

i understand these voltages are out of spec the KOEO i feel may be OK but the idle is low

that said the pressure values indicated by an actron CP9580A and OTC Genisys fall within acceptable range

manifold vacuum measured at sources solenoid is approx 20" at idle

MAP sensor tracks while driving

TPS tracks properly from 0% to 100%

 

my friend worked for Subaru for years and is a skilled tech (as am i but not experienced with subaru) he hasn't personally checked the car i've been bouncing stuff off him and he's stumped as well

 

two points to note, i do not have the ability to do the "idle switch signal" test though i "assume" because the TPS tracks from 0-100 that should be OK

 

second is that with either scanner hooked up when the key is cycled on neither gives a MAP value it just shows "0" though the other values come up and TPS will track. when the engine is started the MAP will show an atmospheric value of approx 30.1" (102kPa) , then when the throttle is bumped and the solenoid switches it will show a manifold value of 8.5" (29kPa). i'm assuming this is how the system is supposed to work where it reads atmospheric until the TPS is bumped though i'm not certain. the fact that the MAP shows no value with just KOEO concerns me i would "think" it should read but of course it does read once the engine is started.

 

a new MAP sensor is $300 and has to be ordered, i'm not going to "try" a new one at that cost. ordered another used one off ebag will try that though i'm not very convinced the sensor was ever an issue. after all the initial tests and swapping the sensor my first impression was that the ECU was the issue but i just let it go because the car ran Ok and initial checkes showed no problems. now i'm down to trying another MAP and if that doesen't work will have to try an ECU

 

what am i missing?

 

thanks and sorry for the long winded first post

[idosubaru]

any engine work in the past year?

 

vacuum hoses - you're positive there's no leaks?  have you checked the brake booster hose?

 

those sensors never fail, you would be wasting money on a new sensor if you've tried three others.

 

you've hit some extensive areas - the FSM seems to indicate testing the pressure source solenoid valve and associated vaccuum hoses, i'd walk through the FSM to diagnose something that you've spent so much time on, diagram from the FSM for that valve is here:

http://www.subaruoutback.org/forums/66-problems-maintenance/11297-cel-p0106.html

[jamesc]

no engine work, hose routing is simple and correct. i've been referencing Alldata, Mitchell and FSM information which best i can tell is what Alldata has. i did bench check the sources valve and found it to be OK. what's throwing me off is that unless i'm screwing something up the absolute pressure values indicated with the scanner fall within range

[johnceggleston]

could it be a wiring issue?

mice, dirt, loose connector?

 

 

 

A P0106 could be caused by:

Bad MAP sensor
Water/dirt intrusion affecting MAP sensor connector
Intermittent open in the reference, ground, or signal wire for the MAP sensor
Intermittent short in the reference, ground, or signal wire for the MAP sensor
Ground problem due to corrosion causing intermittent signal problem
A break in the flexible air intake duct between the MAF and the intake manifold
Bad PCM (do not assume the PCM is bad until you've exhausted all other possibilities)

Read more at: http://www.obd-codes.com/p0106
Copyright © OBD-Codes.com

 

Edited December 27, 2013 by johnceggleston

[jamesc]

can't say for certain that it's not an intermittent problem but everything has been inspected pretty thoroughly and i've never seen a problem while driving with the scanner hooked up

[idosubaru]

the ECU's never fail, Subaru ECU failure is almost nonexistent.  what that means is buying a used one is cheap because they're not worth anything and a good bet. they're also really easy to replace and might be worth a try.  i know of one ocassion where ECU replacement did fix an oddball issue - i just can't recall what the issue was.

 

post in the parts wanted forum here.  i've got a few EJ25D ECU's but i don't know if the 99 may be an odd ball year due to a few changes in 99.  someone will have one or www.car-part.com

Edited December 27, 2013 by grossgary

[jamesc]

update,

 

my friend has an ancient subaru select scanner but didn't have the exact
module for this car. we attempted to use it in an effort to see of the
idle "switch" indication was correct but couldn't get the scanner to
work.

i'm inclined to think this isn't an issue though because the TPS values
are good at 0-100 and the source solenoid switches to manifold as soon
as the throttle is bumped while running.

because the freeze frame data shows a value of 30.1" when the code sets
while the engine is running i decided to try running the manifold
directly to the sensor. i erased the P0106, rerouted the hose then drove
the car. it is NOW setting code P1143 (and only this one) which is
sensor performance low. i'm going in circles because this new code makes
me think that the ECU WANTS to see atmospheric value when first started
until the throttle is opened but yet when the system is correctly
routed it's setting the code because it's 30.1"

i can now duplicate the CEL condition. after erased and everything
properly routed on the very first start it will set P0106 pending and
then without driving it a few subsequent starts will set the P0106 as a
hard code.

the MAP does track correctly once the solenoid shifts to the manifold

though i haven't perfectly followed the diagnostic procedures i think
i've been pretty thorough and can't help but think this ECU is brain
damaged

this thing is driving me nuts
        

        

        

        
        

        

        

        

    
        


        
        
        
        
         
    
    
        
        
            
            

Edited December 27, 2013 by jamesc

[Fairtax4me]

Seems to me that its setting the code when the sensor is at atmospheric pressure because the ECU is expecting to see manifold pressure at that time.

 

The other code (input low) is being set when the MAP is connected straight to manifold vacuum because the ECU is looking for MAP pressure to rise when the PSSSV is commanded closed (or open, whichever way it works).

 

How did you test the solenoid? Did you blow air through it to make sure it seals properly, and is not clogged? Easiest way to test these is to connect the green plugs under the dash to put the ECU in Test Mode. This cycles all of the solenoids and relays when the key is turned On. Make sure the solenoid cycles and blow air through the different ports to make sure the valve seals correctly.

 

Some have a little filter on one port, others have a line to the Evap system to measure fuel tank pressure. Its hard to tell if the filters clog, If yours has the filter try replacing it. Used should be fine.

[idosubaru]

Some have a little filter on one port, others have a line to the Evap system to measure fuel tank pressure. Its hard to tell if the filters clog, If yours has the filter try replacing it. Used should be fine.

 

if you mean the small quarter sized filter in line with a vacuum hose - i had that filter cause a check engine light on a 1997 Impreza EJ22.  i tracked a bunch of stuff, replaced EGR, solenoids, and had that code for months or a year and couldn't get rid of it.  i think someone mentioned it here and so i replaced it and the check engine light never came on again. i think it was an EGR related code on mine but that was a MAF car and i've never really seen or heard of anyone having that happen.

Edited December 28, 2013 by grossgary

[Fairtax4me]

No, but that one causes problems too. The one I'm talking about is a black cylinder mounted directly on the pressure source solenoid.

[jamesc]

thanks guys i appreciate any and all help. this thing really has me puzzled not to mention i have a 632" all aluminum race engine i'm helping a friend with that's giving me grief

 

it's tough trying to fix things over the phone, i'll try to recap a little

 

>clear code

>connect and setup scanner

>turn key on view data, everything shows values but the map reads zero

>start engine, map reads 30.1" (atmospheric) and immediately sets pending code P0106, don't know how long it will stay there at 30.1 but i let it idle a bit before bumping throttle

>bump throttle, the source valve immediately switches and MAP drops to 8.5"

>if i restart a few times it will set the code hard without even driving and when it hard sets the freeze frame shows the MAP at 30.1

 

this is what was making me think the ECU didn't want to see the 30.1 for any length of time after starting which is why i figured as a test i would route manifold direct. this way i figured it wouldn't set a code until it cycled the source valve to look at atmospheric again which would set a different code. i don't know what the cycle rate is but at sometime after starting i would think it needs to occasionally look at atmospheric.

 

so having done that it really threw me a curve when it immediately set the P1143 code (low value) when i started the engine. so now it's pissed if it sees either 30.1 or 8.5

 

i know i have to be missing something simple here but it seems counter intuitive. it triggers a code both low or high depending on how the vacuum is routed i can't get it to not trigger some code.

 

i checked the source valve though i didn't use air pressure or vacuum pump i just blew or sucked on the hose (yea i know) and it seemed to shift tight which is kinda proved out by the MAP readings seen when it cycles when started and throttle bumped.

 

i checked the filters, the hoses are new and i feel confident about the system vacuum integrity

 

as mentioned the measured voltage values are off but the indicated MAP values both atmospheric and manifold are within range

 

the "idle switch" is still an unknown though afaik there is no actual switch just the normal potentiometer and the TPS values are solid at 0-100%

 

the fact that the source valve switches as soon as the throttle is bumped leads me to believe the "idle switch" is working correctly

[jamesc]

freeze frame from P0106 code

 

 

 

P1143, sorry bout the big pics i did downsize them