Boost exceeding target?

I recently was retuned and have an 18.5 psi boost target. I'm regularly exceeding that - although pretty much only in 5th and 6th gears. High boost on my logs is 20.72 psi. On stock turbo, fuel pump and injectors. Would anyone mind taking a look over the attached logs and let me know if this is something to be concerned enough about to get the tune adjusted? I had to delete some of the data in the logs to make them small enough to attach, but there should be incidents of exceeding the target boost in the portions attached.

 

I also have the full logs in Excel format - PM me if you want to view those and I can email them direct.

 

Thanks!

datalog22.5.csv

datalog23.5.csv

Next time you post logs, you have to make sure it's formatted correctly. The quotations marks inside the CSV/TXT file mess up Excel's ability to import the data into the correct columns.

 

Anyways, after some fiddling I was able to open it.

 

Mild boost spikes in cold weather aren't a big deal IF you don't knock.

 

You are knocking, so yes, it's something you should be concerned with. In datalog 22.5, looking at row 967, there is FBKC. 2 counts of it back to back. And the reason for this is semi-apparent: your tuner either dialed in a lot of WGDC at 3K, or he messed with the Turbo Dynamics values (you didn't log them, so I can't tell.)

 

It helps with spool in lower gears, but when you're in higher ones.... well... you see the problem.

 

It happens again on line 1373.

 

Suggest you ask your tuner to fix it. I have no idea what your map looks like, and it's an AP anyway, so my ability to help is limited there.

 

Also, your fuel trims look a bit out of wack. YOU need to fix that. Do a quick 2 second log of ONLY the following parameters while you're in neutral and not moving:

A/F Learning 1 A

A/F Learning 1 B

A/F Learning 1 C

A/F Learning 1 D

 

That will tell us what your fuel trims are and if something needs to get fixed.

Thanks, I'll do the quick log on the A/F Learning and post it by this weekend.

Here's the requested log file. Just a few seconds of idle after warm up. I didn't open up the .csv file at all with, so it's in whatever format the AccessPort exports it in.

 

Thanks for the help and feedback!

 

Crap, just checked the file after loading it and see that only 1 of the learning figures actually logged. I had checked through all 4, they showed logging while I was doing it, so I must have done something wrong...

datalog24.csv

You don't have ANY of the parameters lol. It's A/F Learning #1 Range A or something like that. You need Range A, B, C and D.

How about these ones? Yep, they have the right data.

 

First one was still warming up. Second one was after a short drive at a light.

datalog27.csv

datalog28.csv

Okay, yeah, I have no idea what your tuner did (if anything) but your idle fuel trim is way out of wack. Your ECU seems to think it is very rich at idle, so it's pulling fuel like crazy (-25%).

 

What sort of intake do you have?

 

Although that's not really relevant to your boost spike problem.

SPT intake...

Hrm, that could be it. Hard to say. The SPT intake, if you read around here, sucks. And having an aftermarket intake in general throws off your MAF anyway. Again, without looking at your tune, I have no idea what your tuner did. But having your fuel trims at -25% is most definitely not right.

Haven't been on the forum much over the holidays, so didn't see this thread til just now.

 

As iNVAR already mentioned, your idle fuel trim is way out of whack, although the B, C & D look more or less OK. AF Correction appears to go to 0 at WOT so your tune might be set up to not apply any D range corrections at WOT (my tune is set up this way).

 

You're probably best off bouncing your logs off your tuner, but if you'd like, PM me your calibration file and I can see if I can take a look at a few things, namely the WGDC, WGDC IAT Comp and Turbo Dynamics tables (I have AccessTuner Race, my self-modified protune is AP-based). I'd avoid going WOT like you did in your logs in the higher gears until you either get the boost spikes toned down a bit or get some timing pulled from the problem RPM/load regions.

 

What brand/octane gas are you using? If you aren't already doing so, stick w/ major brand stations. I have numerous datalogs from my previous car (stage 2 '05 Saabaru 9-2X Aero aka WRX) showing consistent knock w/ some 2nd tier stations, which was completely 'fixed' by switching back to Shell/Sunoco/etc gasoline. I presume you're running 91 or whatever the max readily available octane is in CA.

 

I've spent the last few months getting my boost dialed in. It was a learning curve that required quite a bit of datalogging in various driving/temperature conditions and some adjustments to WGDC, Turbo Dynamics and WGDC IAT Comp tables, but I'm finally pretty happy. I still hit peak boost w/ my VF52 by 2.9k-3k RPM as I did w/ the original protune, but the boost curve is smoother, almost never overshoots target by more than 1 psi, and doesn't shift by more than +/- 0.5 psi b/n 25F and 80F. I don't see overboost knock anymore, even when mashing it going up mild inclines w/ a passenger and full tank of gas.

Gas wise I always use Costco.

 

Sent the logs to my tuner, talked to some folks @ Cobb and also sent them the logs. We'll see what happens next. Hopefully I can get this resolved sooner rather than later.

 

Thanks for the help!

I wouldn't think that changing the gas brand would have much impact on idle fuel trim, but if it were me, I'd probably run the current tank close to empty and fill up w/ something like Shell, Sunoco, etc. costco gas might be good, I haven't looked for feedback from others who have been running it. That said, many yrs ago I filled up @ a local Valero w/ my stage 2 05 9-2X Aero. The station had been Mobil for ages and I had no problems w/ the Mobil gas. The owner retained control of the station but simply changed the branding. He showed me pages and pages of supplier invoices that revealed Valero was actually the supplier of the gas that was delivered to the Mobil station (so switching to being Valero branded cut out the middle man and associated costs). One night I was datalogging and saw FB knock all over the place at mid-high loads and mid-high RPMs. After a day or 2, that knock ended up in Fine Learning as seen in Live View (open source tool).

 

I burned off that tank, filled up at a Shell station, drove around for a few miles to get the new fuel into the lines, and repeated the datalogging. No knock. I ran that tank down and filled 1/3 to 1/2 tank w/ the Valero again, drove around for a bit, then repeated the datalogging, and all the knock came back. I have never filled up at a Valero station since.

So, on another log file (too large to upload) I'm in 'I' mode. I get one instance of -1.4 knock at about 2300 RPM, negative boost(-3.9), partial throttle(18.04). Is this related or phantom knock or ?

i think there's similar FBKC in your previously posted logs (datalog22.5, datalog23.5).

 

datalog22.5:

-- single count at row 190, at 0.85 g/rev load & 1322 rpm, decays to 0 by row 199

-- single count at row 661, at 1.02 g/rev load & 2410 rpm, decays to 0 by row 666

 

datalog23.5:

-- single count at row 1067, at 0.93 g/rev load & 2860 rpm, decays to 0 by row 1070

 

i don't think it's related to the high load (WOT) knock that's seen elsewhere in your logs.

Okay, I flashed back to my old map. Here's the logs from today. A couple of idle or low speed logs and one with WOT. Let me know what you think... I see a bunch of Feedback Knock, so I'm thinking that's a very bad thing?

datalog36.csv

datalog37.csv

datalog38.csv

datalog39.csv

You are logging so much stuff that your logging interval is OVER half a second, which makes your logs.... not worth a whole lot.

 

And what do you mean you flashed your old map? What old map?

 

P.S. Those logs are impossible to read, and you'd be hardpressed to find someone willing to sift through it. I'll tell you why though:

 

1) Your logging interval sucks, as mentioned above. You need to remove a lot of stuff. All those A/F parameters? Get rid of them, except AF Sensor 1, and AF Learning 1 (Not ABC or D. JUST AF Learning 1). Get rid of accelerator position. You have throttle opening angle! Get rid of your injector duty cycle unless you suspect you're running out of fuel from your injectors being too small (not happening on stage 2.) Get rid of Knock Act Sw. Get rid of MAF Volts (unless you're rescaling your MAF).

 

2) Your log is huge. When you're doing a single pull (small log) at WOT, you know what gear you're doing it in. You just give us a huge driving log. We have NO idea what gear you're in at any point. We can make an educated guess, but it's annoying. If you're going to give us huge logs with lots of driving, put in gear.

 

As an example of why gear is important... in datalog39, right at the beginning, you are applying 30-100% throttle there for a brief amount of time. What gear were you in? It looks like probably 4th or 5th from how slow your engine RPM climbs, but I can't tell. You do it again on line 79, 2700 RPM and 23% throttle. What gear?

 

Also, you have learned knock ALL over based on your log, and a LOT of it so for now, I suggest you

1) Tell us what map you went back to.

2) Talk to your tuner if they're willing to help you. Otherwise, come back here and

3) Stop experimenting with maps until you know for sure what you're doing. I don't want you to hurt your car, and I think neither do you.

Okay, here's the back story:

 

Bought the car new in February 2007. In March 2008, it was tuned at Harman Motive for my AutoSpeed TBD, AVO TMIC, AVO Boost Actuator and SPT Intake. This was at 15,800 miles.

 

May of 2011, 50,800 miles, the car started misfiring. Took it to Subaru - piston #4 had broken it's ringlands. Entire shortblock was replaced as well as the turbo and clutch. Reverted back to stock exhaust and intake and the Cobb Stock Style map.

 

August of 2013, reinstalled the exhaust and intake, took it to be tuned by Shamar of Infamous Performance. Shortly after getting the new tune going, I was freaked out by the high boost I was seeing in 5th and 6th gears. I sent a message to Shamar who said that my boost gauge was off. I was very skeptical about this and went back to my Harman Motive map since that only sees 16 psi of boost. Recently I decided to try and get the issue with the Infamous map resolved. However, having searched the forums and attempting to contact Shamar, it seems he's vanished from the scene, with money from various members.

 

So, now I'm sort of stuck. I have AccessTuner Race access to my Infamous map, but am a complete noob on Subaru tuning and don't want to mess it up. I'm thinking that the safest bet is to do a compression check now and see where things stand. If compression is good across the board, I'm leaning towards taking it direct to Cobb in Fountain Valley to be retuned. If I have an issue with compression, hopefully I'm early enough to fix it with new rings, check the valves, etc. rather than needing another new shortblock.

 

I did as asked and filled up (only had a half tank to fill) with Chevron 91. Since I had a half tank of Costco gas already, I added some Lucas octane booster. I reduced the logged parameters to increase the logging frequency. I've got a couple of hopefully more useful logs here (copies in .csv and .pdf formats attached):

 

Log 1 (datalog52)- First gear to fourth on hard pulls, then coast off on 5th and 6th.

 

Log 2 (datalog53)- 6th gear partial throttle on the freeway at lower RPM level that produces Feedback Knock.

datalog52.csv

datalog52.pdf

datalog53.csv

datalog53.pdf

You're still logging stuff that you shouldn't be, and not logging stuff that you should be. Please review my post and make adjustments accordingly. Of particular note, you should still log GEAR when you're going through multiple gears in a log.

 

A pull is a straight pull in one gear from 2K-redline. Shifting makes things more complicated.

 

As far as your second log goes, here's the solution: don't putt around in 6th gear at under 2K with partial throttle. You're bogging your engine down.

 

You should most definitely get returned. Yeah, since you have a Shamar map, you are now left ass-out. I wasn't happy with my Shamar stage 2 map about 2 years ago and tweaked the hell out of the tune and now I'm happy with it.

 

P.S. SPT intake sucks. Causes problems. Difficult to tune for properly. Go back to stock intake BEFORE you retune.

Also I'd always do compression and a leakdown test. I think sometimes compression can be "ok" while there is excessive leakdown. (Sounds like usually with something like a broken ringland there's both bad compression and leakdown, but I have very limited experience.)

 

I'd take the advice given on this and the other thread.

---> Go back to stock intake, load the Cobb 91 map, check for knock using littleblue's sticky info and take a learning view snapshot, and check compression and leakdown (can buy a leakdown tester (OTC) for about $100 or so).

 

There are better intakes that work fine with stock MAF mapping if you want to put those on later (and some that require MAF scaling change).

Changed back to the stock intake, changed WGDC on the map for 330 requested torque levels as per tuner recommendation. I updated my logging parameters and will make a couple of new logs tomorrow. I'll post them up here tomorrow evening...

 

Thanks a bunch for the feedback.

So, a couple more logs here.

 

With the stock intake on, I went to the Cobb OTS State 2 ACN91 map.

 

Datalog 60 shows WOT runs in 2nd and 3rd - no knock at all.

 

Datalogs 61 & 62 show significant knock - 2nd & 3rd, partial throttle, lower RPM, no boost, high calculated load because I was on hills and in normal traffic.

 

Hopefully I'm logging all the right parameters now?

 

Feedback?

you're likely comparing apples to oranges. knock control is keyed off load & rpm. just bc you get no knock WOT at high load & high rpm doesn't mean you'll get no knock @ high load & low rpm. where are the logs, post them so we can see exactly what you're car is seeing.

Sorry, meant to attach these before. Added 1 more too.

datalog60.csv

datalog61.csv

datalog62.csv

datalog63.csv

Is this the cobb ots map and your stock intake?

Yep. Just loading the tweaked OTS map now.