Subaru Forester Owners Forum banner

1 - 20 of 75 Posts

·
Registered
2003 Forester 2.0 XT Manual
Joined
·
683 Posts
Discussion Starter #1
Posting here as this seems to affect some Forester and Imprezza models with Electronic Throttle Bodies or IACV that undergo a battery disconnect, and lose their idle calibration.

As per these links to retrain the ECU idle parameters:

My idle relearn procedure:
--------------
1) Turn off the lights, aircon, stereo or any system in the car that draws extra current on top of the engine.
2) Disconnect the battery for 30 mins.
3) Reconnect the battery.
4) Before you start the car for the first time, turn the key to the ON position but do NOT turn the engine over. Wait 10-15 seconds so the electronic throttle body or IACV has time to go to the factory programmed home position.
5) After waiting, start the car and let it idle without any load, lights, A/C etc.
6) Every 20 seconds or so the idle will be adjusted up and down as the ECU tries to adjust it towards a stoichiometric fuel / air mix.
7) Leave the engine running for a full 10 mins but DO NOT TOUCH THE ACCELERATOR during this time or turn on anything that will cause extra electrical current draw.
8) Turn off the engine, and leave the key in the OFF position for at least 20 sec.
9) As per step (4) turn the key back to the ON position for 10-15 sec without actually starting the engine.
10) Start the engine and leave to idle for a further 5 minutes without touching the accelerator and without turning on other systems in the car.
11) Turn off the engine again and wait at least 20 sec before restarting.
12) Take the car for a test drive as the ECU should now be fully retrained.
--------------
I know for a fact that this is necessary for the 2003 Forester XT because this process seems to have been the cause of all the teething problems with the LPG system because it was never retrained after the battery was disconnected during the conversion. I used an OBD2 adapter to measure the AFR beforehand - and the LPG AFR was waaay out in comparison to the petrol. After following the process above both sets of AFRs are now very close to a stoichiometric mix, and the car idles correctly without the revs bouncing up and down constantly.
 

·
Registered
2004 XT AT
Joined
·
1,210 Posts
What a load of B.S. Internet mythology. I mean I'm glad it worked for you, but my battery has been disconnected dozens of times with no special "training" afterwards. I have a hard time believing any competently designed car would require a spurious mystic undocumented ritual to reset it after a battery disconnection!

-- Steve
 

·
Registered
2003 Forester 2.0 XT Manual
Joined
·
683 Posts
Discussion Starter #3 (Edited)
1) Calling BS just because you havent needed to do this yourself isnt very helpful and is a good way to put people off posting up hints and tips in future.

2) Its common knowledge that certain ECU parameters are lost when the ECU is reset (fuel trim corrections etc) so its not all that unreasonable that the idle position is also lost given that the ECU doesnt necessarily know in advance whether its running on 95 RON, 97 RON or 99 RON in order to adjust the fuelling.

3) One of the linked posts is to this very forum (albeit a different area) - with several other SubaruForester.org posts outlining the same process of retraining the ECU as the solution of idling problems - and people commenting that it worked for them. Some people even quoted that the process is outlined in their manual - but its not in my UK spec manual.

4) I was initially sceptical but decided to give it a try because THIS POST refers to the ECU having to send a signal to the injectors to 'catch' a rapidly falling rpm level (e.g. dipping the clutch) and if the ECU doesnt send this signal in time then the engine can drop to low rpm and be on the verge of stalling for a few seconds. These were the exact symptoms I was having on both petrol and LPG.

5) Petrol seems less susceptible to the problem than LPG - but that doesnt mean the problem isnt occurring on petrol. As per comment #4 dipping the clutch and watching the point at which the needle stops falling is a good way to see whether you have a problem. If the needle dips to 500rpm or lower then the ECU isnt sending the injector pulses quickly enough to hold the revs at the normal idle level.

6) If you miss the symptoms in the first week or so after a battery disconnect then the idle correction still keeps learning - as does the long term trim correction. As a result the car will usually sort out the idle problems after a week of running.

7) Before I trained the idle the AFRs were off - with both petrol and LPG not getting to a stoichiometric mix very quickly after letting the revs drop to idle. After training then on both fuels the AFR drops immediately to 14.5:1 to 14.8:1 which is almost bang on an ideal stoich mix.

8) During the process you can actually see the ECU raising and lowering the idle slightly every few seconds as it hunts for the optimum level. On subsequent engine restarts the engine dropped immediately to the same idle level as the end of the training - lending credence to the fact the ECU is actively learning the idle position.

Given that this process takes very little time and effort then I think its worth trying if you have already ruled out MAF, O2 sensor and vacuum leaks as the likely causes of any idle problems.
 

·
Registered
05 XT2.5
Joined
·
1,999 Posts
Sorry but I'm in the BS camp here too. Well, not BS but way overcomplicated. I've used the Vishnu method dozens of times.

Turn off everything. Disconnect the earth strap. Hit the brake pedal to use any residual current. Wait between 5 seconds and 2 days. Reconnect the battery. Drive without using boost to a road that is preferably long and straight and slightly uphill. Start to accelerate allowing boost to rise slowly and without getting too much. Continue for 30 seconds or thereabouts. The end. Your car will now idle and accelerate more smoothly and feel a little more powerful. If you haven't done this for a long time or even ever the effect will be more remarkable.
 

·
Registered
2004 XT AT
Joined
·
1,210 Posts
Sorry about the tone of my earlier reply. Like I said, glad it worked for you, but I am pretty sure that the ECU will perform what little learning it does during normal driving. Maybe LPG changes that for some reason.

-- Steve
 

·
Registered
2003 forester 2.0 xt at
Joined
·
699 Posts
Actually simuilar method for my old honda also, changed batteri, and car did not want to run at idle. Changed a whole lot of parts, then i found something like this! Guess what? It worked! So it just needed a "learn idle prosedyre" :)
 

·
Registered
2004 2.0 XT Manual
Joined
·
230 Posts
thanks for this bod, im going to give it a try when i get the chance. i changed my battery recently and didn't do this.

had to do a similar think with the golf after i cleaned out the idle valve
 

·
Registered
2008 GRB 20th anniversary 6 MT
Joined
·
6,529 Posts
You don't need to do this; also there is no need to disconnect the battery for 30 mins. My AM (so I guess all similar eraseable data) go back to default as soon as the battery is disconnected.

If the OP encountered a problem that was due to the fact that the petrol ECU didn't have a chance to recalibrate hence the LPG ECU was calibrated using the wrong baseline.

If you don't use LPG it will take as long as it takes for the petrol ECU to recalibrate, there is no big deal (similarly with the AM going back to 1)
 

·
Registered
2006 STI
Joined
·
6,953 Posts
You don't need to do this.

If the OP encountered a problem that was due to the fact that the petrol ECU didn't have a chance to recalibrate hence the LPG ECU was calibrated using the wrong baseline.

If you don't use LPG it will take as long as it takes for the petrol ECU to recalibrate, there is no big deal (similarly with the AM going back to 1)
So it may well help Tumble then?
 

·
Registered
2008 GRB 20th anniversary 6 MT
Joined
·
6,529 Posts
Yep, I meant to write you don't need to do this if you don't use LPG.
 

·
Registered
2004 2.0 XT Manual
Joined
·
230 Posts
good stuff :)

should i do the whole process on petrol, on gas or both?

im guessing petrol to create the correct baseline.

about the time my stalling happened was when my old battery was failing, so i presume every time it was dead i would end up with the wrong idle point.

once sorted id of thought i could get away with taking off the low end lean - which i think is causing the odd engine surge i mentioned in bods other post when warming down on gas. (checked today and it dosn't happen on petrol)
 

·
Registered
2003 Forester 2.0 XT Manual
Joined
·
683 Posts
Discussion Starter #13
Yep, I meant to write you don't need to do this if you don't use LPG.
Technically you do :icon_razz:

I found earlier today that running an ECU Fault Code reset using the Torque Pro application via OBD2 adapter actually triggers a FULL ECU reset, which clears out the codes but also the trim correction data, and yes - you've guessed it, the learned idle position too.

If I drove the car on ONLY petrol immediately after the reset (with no learned idle) then the original problem of low idle (bordering on stalling) returned immediately, and didnt get any better after 15-20 mins driving.

I then did another reset and this time did the ECU learning on petrol alone. After only 5-6 mins of stationary idling with everything turned off except the engine it had settled down to a steady idle. I then took it for a quick spin and it was running perfectly fine without fault.

I then reset again - and the idle fault returned immediately.

Finally I did another reset and did another idle learning session - this time swapping between petrol and LPG during the session. It took slightly longer before the idle was fully settled - but once it had the AFRs and Fuel Trim Corrections (both short and long term) were settling extremely quickly. In fact swapping between LPG and petrol the ECU was managing to move the long term trim correction average within 90 seconds of changeover, which is pretty damned good in my opinion. All idle AFRs were also VERY close to a stoich mix on both petrol and LPG.

:woohoo:

So as far as Im concerned its worth performing on any car showing a rough idle. Im still investigating the cause of my idle problem, and will check the vacuum hoses and throttle body at the weekend to see if there is an underlying problem thats throwing the default idle position off slightly.

HOWEVER - as the car was originally LPG mapped with the idle fault in place, and the AFRs all over the place then I will have to go back for a remap tweak - especially as Im still getting occasional P0171 codes indicating its running slightly lean with low throttle (perversely its fine on WOT!). I have also configured Torque to show the Fuel Trim corrections so I will be monitoring them when the CEL next appears to see if its the short term or long term figure thats raising the fault code.
 

·
Registered
2010 Forester Touring 4 speed Automatic
Joined
·
43 Posts
Replaced my car battery 1 week ago and my fuel consumption is over the chart (21litres per 100km) usually would be around 10-12 litres. So i saw this thread and wondered if I should do it. So I call my SUbaru dealer and he told me that I had to do the idle relearn with my 2010 Subaru Forester. So I will give this a try for sure.

Thanks BOD.
 

·
Registered
2010 Forester Touring 4 speed Automatic
Joined
·
43 Posts
well since yesterday gas consumption is still at 25l/100km (9.41 mpg) so there still seems to be someting wrong. Wonder if I busted something when I boosted someone the other day at the train station....
 

·
Registered
2009 Forester Limited Manual
Joined
·
117 Posts
Just adding my $0.02 to the conversation of CEL: P0171.

My 2009 popped a P0171 a couple times over the last couple weeks - and the car would rough idle and almost die when coming to a stop at a red light.

I did not have any CRC MAF cleaner, so I tried the Idle Re-Learn steps with no luck.

Upon getting some CRC MAF cleaner, I removed my MAF and found a piece of dandelion fuzz stuck to the MAF 'sensor wire thingies'. I removed the fuzz & spray cleaned the MAF.

The forester runs great now, no CEL, no rough idle.

Thanks
G
 

·
Registered
2012 Forester FXT automatic
Joined
·
35 Posts
Thanks for the reference link to my post BOD. For those claiming BS... say what you want, you're wrong. The process I posted solved the stalling problem I was having in my MT 2005 Forester after I had run the battery dead. I ran the battery dead because I had broken the alt belt and was trying to get the Forester back to civilization. For about a week it stalled whenever it idled, including when I was downshifting on an exit ramp going 50 mph. After I did the "relearn" process the problem went away. I did nothing else, so this process DID fix it.
 

·
Registered
05 XT2.5
Joined
·
1,999 Posts
Just adding my $0.02 to the conversation of CEL: P0171.

My 2009 popped a P0171 a couple times over the last couple weeks - and the car would rough idle and almost die when coming to a stop at a red light.

I did not have any CRC MAF cleaner, so I tried the Idle Re-Learn steps with no luck.

Upon getting some CRC MAF cleaner, I removed my MAF and found a piece of dandelion fuzz stuck to the MAF 'sensor wire thingies'. I removed the fuzz & spray cleaned the MAF.

The forester runs great now, no CEL, no rough idle.

Thanks
G
Glad it's better but how did dandelion seed get onto your MAF?
 
1 - 20 of 75 Posts
Top