2008 Spec.B makes noticeably more power after battery disconnect.

[SilentJ20]

Not sure where to start investigating this.

As per the title, is that due to anything troubling?  I'd expect as the ECU re-learns it would change some.  But the first time I did electrical work and disconnected the battery, then reconnected and drove it was a serious difference.  The ol' butt-dyno was pushed back in the seat much harder.  18mpg around town (not too far out of expected).

Then, as time goes on, the power faded over the course of a couple days.  No CEL.  Also, at WOT the power 'stutters' a little bit, like it's trying to figure out how much power to make.  And occasionally stumbles hard at engine free-spinning back to idle (almost dies).

I've heard some or all of these could be due to intake plenum seals (during warm-up they may let in un-metered air at various temps), or the forward O2 sensor, or the throttle body TPS...

Searching the forum regularly or with google didn't really help with the battery reconnect power boost thing.  Not sure if I have one problem or many.  Car drives fine for the most part.  Any thoughts?

[m sprank]

Chances are your IAM was less then 1.0.  The ECM reset after the disconnect. If the problem persists, the car will lose power as the IAM drops below 1.0.  No learning view, no data.  Its all hypothetical.  But from your decription, this sounds more than plausible.  Why does the IAM drop below 1.0?  Normally, knock. 

[SilentJ20]

This makes sense.  It does kind of resemble behavior o my previous daily when my sister-in-law filled it with Reg instead of Premium.  I've suspected (and later confirmed) that my engine had a tune due to the COBB intake.  But, I have no AccessPort or any other indication as to the type/maker of the tune.  I guess my next step is to try and get Rom Raider up and running, learn about Learning View and get logging.

[SilentJ20]

Installed LearningView and took a screenshot of what the ECU has been dealing with.

My IAM is hosed, some knock-detected at high load/rpms, and my trims are all over the place.

I'm struggling getting RomRaider installed to get better logs, but I have some suspicions.  These are indications of running lean, correct?  Is this something that could be caused by a previous owner installing an aftermarket intake (COBB) and not getting a tune done?  I know that messes with the MAF accuracy.

[m sprank]

Ok.  The car is dorked bad.  IAM is ZERO and you still have knock.  So, if the tune asks for 10 degrees of advanced timing you multiply 10 by zero and this is your timing value.  But, you still have knock.  The ECM can not adjust anymore to alleviate the knock.  You have a lot too.  The emc is pulling a shit ton of fuel at idle and adding a shit ton of fuel at WOT.  Technically the +15.6 is beyond the ecms 15% adjustment range.  You are rich at idle and lean at WOT. 

 

This could be a combo.  Not a great tune and physical issues under the hood.  The car is not very safe to drive at this time as engine damage is very possible.  Things to check, MAF, O2, smoke test for leaks, loose/rattling equipment almost anywhere on the car, fuel quality, tune.  A basic log at idle would be better than nothing to start with.

[Pleides]

I'd have a speed shop boost leak test that car before driving it. Where are you located?

[SilentJ20]

Western Washington.  There's a good shop nearby, PIA, that I'll try to get it into soon. Usually I'm a diy guy. But if this ends up needing professional help I understand.

[Pleides]

18 minutes ago, SilentJ20 said:

Western Washington.  There's a good shop nearby, PIA, that I'll try to get it into soon. Usually I'm a diy guy. But if this ends up needing professional help I understand.

They might be able to lend you a working MAF sensor or another member can let you borrow one to see if that helps with your issues. I had very similar issues to you when my first MAF when bad. Use only OEM Denso sensors. Aftermarket ones are an absolute no-go on a modified car, or really any turbocharged car.

[SilentJ20]

I'll be getting an idle log soon, but I want to clarify one thing.  This car has an aftermarket COBB intake, THIS ONE.

IF whoever installed it did not get the ECU tuned as per COBBs instruction (because the MAF readins need to be recalibrated), is it possible this is a major player in my problems.  Reason I'm asking, I can replace MAF, O2 and lots of other things, but if the intake doesn't work with the tune, then I'm chasing the wrong symptoms.

Is there any way to ID what kind of tune I have?

Idle log coming soon.  I hope...

[SilentJ20]

Got a basic idle log done (I think):

Spreadsheet

My first log, ever.  So it may not have all the relevant parameters.  I didn't run it very long, but the car was at a steady state of idle.  The fact that IAM reads 0.0000 makes me think maybe the knock detection isn't being correctly captured.  And chance a tune moved the ID to another location?  Should I try an ECU reset to see if it starts at optimal and then backs down?

Should I keep this discussion here, or would it be best to move it to a more relevant section?

[seanyb505]

This section is fine there's a lot more traffic anyway.

I've not owned a repair shop before, so take this as a random internet commenter.. 

Not sure I saw this mentioned are you still running the Cobb intake with it as yet undetermined state of tune? I might be inclined to get a stock intake back on the car and reset ecu. 

There's a quick and dirty leak test you can do, unplug the small hose from the bpv and blow into it. Once you've got some lung power into it and you can't blow anymore, plug it with your thumb. Wait 5-10 seconds and let go. You should feel a small rush of air escaping. It's not fool proof, but like I said it's quick and could illuminate larger sizes issues. 

[m sprank]

The intake will throw off the maf reading as it needs to be rescaled for the inlet tube piping size.  This WILL NOT BRING YOUR IAM TO 0. No one adulterated the knock sensor signal in the tune. No one made it more sensitive.  I will look at the log when I get to a pc. 

LO mofo loud at blowing in a vacuum line.  I charged $50 for a smoke test at 14psi.  Dont waste your time chasing ghosts. 

 

We can play all day on line. If you are not knowledgable, skilled and with the proper tools you will be changing parts in an attempt to change the right one. Like winnning the lottery.  Why not just shoot yourself?  Bring the car to a competant Subie mechanic who has acces to a good tuner. Together they should have it worked out. 

[m sprank]

@SilentJ20Need to see MAF g/s and MAF volts to determine if adjustments have been made for the intake. 

[seanyb505]

Meh

I prefaced it with my qualifications. Also didn't advise firing a parts canon to fix a problem. Most shops present themselves as competent but don't really have the passion you have to follow through to solutions. Just sharing what a reputable tuner posted in hopes it adds to the pool of info. 

 

[Pleides]

5 hours ago, seanyb505 said:

unplug the small hose from the bpv and blow into it. Once you've got some lung power into it and you can't blow anymore, plug it with your thumb. Wait 5-10 seconds and let go. You should feel a small rush of air escaping. It's not fool proof, but like I said it's quick and could illuminate larger sizes issues. 

You really need to vacuum test these at like 1 bar because of the turbo. Breath ain't gonna get you there.

[blackobxt]

Leak test for sure. Check all lines especially the one to the fpr. No positive pressure on the fpr in boost equals no increase in fp

Edited February 18, 2023 by blackobxt

[SilentJ20]

Did another idle log with the MAF values added, and some other stuff (more rows on full spreadsheet):

NOTE: this is on the same spreadsheet link as above on the 'Idle_MAF' tab at the bottom

 

Also, noticed when I connect my ECU to RR, I have no Front O2 sensor option.  May have been deleted.  Going to look into that next.

Also, is there any way to make RR save parameter lists between uses?  I can't seem to make it stick...  Still learning.

 

Thanks for all the discussion so far, one thing about forums I like is the community experience and knowledge base.

[m sprank]

On 2/18/2023 at 11:44 AM, seanyb505 said:

Meh

I prefaced it with my qualifications. Also didn't advise firing a parts canon to fix a problem. Most shops present themselves as competent but don't really have the passion you have to follow through to solutions. Just sharing what a reputable tuner posted in hopes it adds to the pool of info. 

 

 

 

The front O2 sensor is all you need to read.  The rear is used for finite adjustments if anything. It is more for emissions than anything. Being brief in this description.  

 

I always had to check my parameter boxes every time I connected. 

Edited February 21, 2023 by m sprank

[m sprank]

At least at idle maf volts appears normal.  Iirc 1.32v = 5.2gs, this is approximate as all cars are a little different.  65mm stock intake tube diam. I want to say Cobb is 70mm. WOT would be the normal next step, but please do not do that. Really need to check for leaks, test front O2 and maf, check tune etc. 

[silverton]

In my experience idle g/s should be really close to engine displacement, unless the COBB intake changes this?

[m sprank]

"idle" is not static.  As the coolant temp is 201 it would appear the car is operating at around 750-900rpm.  The vacuum reading is also indicative of that rpm range and in decent health.  Somewhere between -9.5 and -10.0 would be a good vacuum reading.  The scaling appears close at this rpm.  y = mx + b.

[m sprank]

Dug around and found this.  It is the stock intake scaling from my 08 SpecB:

Stock

Maf Volts 0.9 0.94 0.98 1.02 1.05 1.09 1.13 1.17 1.21 1.25 1.29 1.33 1.37 1.41 1.48 1.56 1.64 1.72 1.8 1.88

Maf g/sec 1.13 1.3 1.49 1.68 1.89 2.12 2.39 2.69 3.02 3.37 3.76 4.17 4.61 5.12 6.22 7.42 8.74 10.22 11.82 13.57

[SilentJ20]

16 minutes ago, m sprank said:

Dug around and found this.  It is the stock intake scaling from my 08 SpecB:

Stock

Maf Volts 0.9 0.94 0.98 1.02 1.05 1.09 1.13 1.17 1.21 1.25 1.29 1.33 1.37 1.41 1.48 1.56 1.64 1.72 1.8 1.88

Maf g/sec 1.13 1.3 1.49 1.68 1.89 2.12 2.39 2.69 3.02 3.37 3.76 4.17 4.61 5.12 6.22 7.42 8.74 10.22 11.82 13.57

Looks like mine's in the ballpark, so at least at idle it's not so far out of whack to explain there is so much fuel being added.  The difference in areas is about 14% from the 65mm and 70mm intake tubes (assuming those are accurate).  And I'm not off that much on a purely MAF reading.

I'm borrowing a leak test setup from a buddy today, so I hope to be able to give that a go in a couple days and see if it brings anything to light.

I do have an appointment at a very reputable Subaru shop (PIA, Pacific Import Auto) on the 28th for diagnostics/investigation if I don't figure things out before that.

[m sprank]

Yes, in the ball park.  All cars run a little different, even stock.  That is why no one generally posts scaling, because they dont want to dork someone else car.  I dont mean for people to input numbers I show as they might not work for that car.  Your idle reading and vacuum look decent. 

Would be rare, but could be noise.  LWFW, bad endlinks, something that taps/rattles or makes a bad harmonic.  Heck, the knock sensor could be loose.  It can be difficult to diagnose cars online.  Good Luck.  If you have questions there are members here trying to help. 

[SilentJ20]

2 minutes ago, m sprank said:

Yes, in the ball park.  All cars run a little different, even stock.  That is why no one generally posts scaling, because they dont want to dork someone else car.  I dont mean for people to input numbers I show as they might not work for that car.  Your idle reading and vacuum look decent. 

Would be rare, but could be noise.  LWFW, bad endlinks, something that taps/rattles or makes a bad harmonic.  Heck, the knock sensor could be loose.  It can be difficult to diagnose cars online.  Good Luck.  If you have questions there are members here trying to help. 

Don't worry, the last thing I want to do is start entering random numbers into my ecu tune.  That will undoubtedly cause more issues.  

By LWFW, do you mean lightweight flywheel?  Just getting my acronym straight.  There is a clunk in the suspension (swaybar bushings I believe) that I will be fixing soon. Is the knock sensor accessible?  I tried google-searching before and, while I know what it looks like, it sounds like it's under the intercooler?  Not to bash subaru, but the fact that it's possible for loose suspension components to cause engine issues has me perplexed.  I understand why you mention it, but just odd.