Data Logs Post Inlet Pipe Upgrade

[cody]

Cory at KSpeed did an excellent job installing a new Perrin Inlet Pipe in my car on Friday. He reset the ECU after the install and I have driven about 75 miles since (to Kingvale, CA). I just did some pulls while logging and I'd appreciate any input. I'll upload them to this post. Even if you only halfway know what you're talking about, you probably know more than I do when it comes to reading logs. If you don't have Excel or Open Office, Google Docs should be able to open these.

Here they are zipped up. I'd start by looking at the oldest ones (sort by filename). You'll note that a couple show me doing a 1st-3rd pull but the rest are all 3rd gear pulls.

[cody]

Mods that Count:

Cobb Protuned by Ed @ Equilibrium Tuning (For two separate elevations, sea level and 5K feet [no map switching necessary]), Crucial UP & DP, Borla Hush CBE, Crucial High-Flow Midpipe, Heat Wrapped Stock Headers, Hallman Pro RX MBC, One Step Colder Denso Iridium Plugs, GP Moto IC Hoses, Perrin Turbo Inlet Pipe, ACT Streetlite Flywheel

Oh, BTW, these were taken at 5K feet of elevation and it's about 90* out. TIA for any good advice.

[sperry]

Okay, I'm finally taking a look at these...

First off, I should have remembered to mention to log the throttle position (TPS). I'm just going to assume these logs are at 100% throttle, i.e. if the rpm is going up, you're flooring it.

Now, looking at the 2nd file (ecuExplorer [AF424-7066] 18-07-2010 12h27m39s, 3rd gear pull, right?):

First thing that's jumping out at me: there's a ton of knock correction in the logs. Like 8 deg around 2500-3000 rpm right as the boost is coming on, but then 6 or 7 degrees of correction all the way to red-line. I dunno too much about tuning the stock ECU, but on my hydra I've only seen like 1 or 2 degrees of knock correction and that's just when bouncing off the rev limiter which triggers the knock sensor. Then again the hydra is a much less sophisticated ECU when compared to the stock ECU... I don't know how much correction you should expect to see... but my gut tells me, that much correction is way too much.

Also, I find it odd that your ignition timing takes a really big dip between 2700 and 5000 rpm. Maybe someone with some real tuning experience can chime in about that... but I would expect to see more timing at low rpm while off boost, then less at high rpm under boost. i.e. I would expect to see the timing drop off at 2700 rpm like it does, but not ramp back up to 30 deg between 5000 and 7000 rpm.

Anyway, like I said, I'm no expert... all my experience is with my Hydra, and a tiny bit of screwing around in open source on the '02 WRX ECU. And everything I've done has involved a wideband AFR gauge... so I feel pretty lost w/o seeing the real-time AFR. I'd get the car over to someone that knows more than me and have them log it w/ a wideband. I'd offer mine, but it's hard-wired into my car and would require you to have an empty O2 sensor bung in your exhaust.

In the meantime, try to stay out of boost, and try not to floor it off boost... since that's where all the heavy knock correction is occurring. And keep in mind, I may be totally wrong about the expected values for knock correction on your car... I'd just err on the side of safety until you can get someone that knows what they're talking about to read the logs.

[cody]

Thanks Scott. Yep, that is a 3rd gear pull ...and all pulls were at WOT.

One cool night, after adding the MBC to my car, I was doing some remote tuning with Ed and I sent him this log (attached) in which I was boosting to almost 18.5 PSI.

He replied with an email entitled "whoa" and this attachment (image attached).

He said my tune looked good though, and as you can see in the log, I was pushing similar knock corrections then. I think it's bad when there are negative values in that column, though I really am not sure.

I think the timing dip is there too, on the old log, and I assume I'm seeing less timing today due to the heat...but I'm just talking out of my ass now.

[knucklesplitter]

On a Subaru ECU the total timing is the "knock correction timing" added to the "base timing". Knock correction timing at 8 is not a bad thing. This means that EQTuning set the knock correction timing to 8 at that load/rpm and all is fine (as far as I can tell). If the advance multiplier is at 16 for 16bit ECU or 1 for 32bit ECU then the ECU will add all the knock correction timing to the base timing. If knock is detected the ECU can pull up to 8 degrees of timing at that point.

To get a quick snapshot of the knock-health while logging, log "IAM" (ignition advance multiplier). It should be at 16 all the time on your car (or 1 on a 32-bit ECU). If it drops then the ECU is seeing significant knock. And if knock drops it far enough the ECU will pull boost too (but not on your MBC=fail car). The couple logs I looked at look fine though. The only thing that would change with a new turbo inlet would be boost response (unless there was a leak), and it is doubtful that the ECU couldn't compensate for that minor change... oh that's right you have a MBC, so nevermind. I would like to see AFR and IAM logged too.

The midrange timing dip is normal. Timing is proportional to load and RPM. As boost comes on down low, timing will drop and then start to come back up as the engine revs and that small turbo's boost falls off.

BTW that Perrin inlet was missing a couple parts. The aluminum BPV recirc. line nipple was missing completely from the kit, so we scrounged one up. Cory also needed some line instead of a tee Perrin supplied. I think that inlet kit is better suited to an STI than a WRX, but either way it should have had that nipple. If it weren't for those issues Cory would have been able to install it in less than the 2 hours quoted.

Next time maybe kick in for a extra half-hour to have me log it for you

[knucklesplitter]

Originally Posted by sperry

Okay, I'm finally taking a look at these...

First off, I should have remembered to mention to log the throttle position (TPS). I'm just going to assume these logs are at 100% throttle, i.e. if the rpm is going up, you're flooring it.

Now, looking at the 2nd file (ecuExplorer [AF424-7066] 18-07-2010 12h27m39s, 3rd gear pull, right?):

First thing that's jumping out at me: there's a ton of knock correction in the logs. Like 8 deg around 2500-3000 rpm right as the boost is coming on, but then 6 or 7 degrees of correction all the way to red-line. I dunno too much about tuning the stock ECU, but on my hydra I've only seen like 1 or 2 degrees of knock correction and that's just when bouncing off the rev limiter which triggers the knock sensor. Then again the hydra is a much less sophisticated ECU when compared to the stock ECU... I don't know how much correction you should expect to see... but my gut tells me, that much correction is way too much.

Also, I find it odd that your ignition timing takes a really big dip between 2700 and 5000 rpm. Maybe someone with some real tuning experience can chime in about that... but I would expect to see more timing at low rpm while off boost, then less at high rpm under boost. i.e. I would expect to see the timing drop off at 2700 rpm like it does, but not ramp back up to 30 deg between 5000 and 7000 rpm.

Anyway, like I said, I'm no expert... all my experience is with my Hydra, and a tiny bit of screwing around in open source on the '02 WRX ECU. And everything I've done has involved a wideband AFR gauge... so I feel pretty lost w/o seeing the real-time AFR. I'd get the car over to someone that knows more than me and have them log it w/ a wideband. I'd offer mine, but it's hard-wired into my car and would require you to have an empty O2 sensor bung in your exhaust.

In the meantime, try to stay out of boost, and try not to floor it off boost... since that's where all the heavy knock correction is occurring. And keep in mind, I may be totally wrong about the expected values for knock correction on your car... I'd just err on the side of safety until you can get someone that knows what they're talking about to read the logs.

See above. You are totally off-base about the knock correction.

Also you can tell a lot by logging the stock O2 sensor AFR along with a number of other things. A wideband would be best but not absolutely necessary, and certainly better than nothing. The stock sensor is okay down to 11 or 11.5:1 AFR then it is useless. If the car goes dangerously lean (>11.5:1) then it will show up.

[cody]

Matt, is there something else that I should be logging? I've got a request from a remote tuner I know (Jeremy AKA Testes1010) to run this software and I can do another pull with ecuexplorer too if I need to turn off a couple parameters and turn others on. Just let me know what I can turn off and what I should turn on.

That sucks that the Perrin kit was weak. It looked extensive to the untrained eye. I didn't mind paying 2 hours labor as it was money well spent.

[sperry]

Originally Posted by knucklesplitter

See above. You are totally off-base about the knock correction.

Also you can tell a lot by logging the stock O2 sensor AFR along with a number of other things. A wideband would be best but not absolutely necessary, and certainly better than nothing. The stock sensor is okay down to 11 or 11.5:1 AFR then it is useless. If the car goes dangerously lean (>11.5:1) then it will show up.

Well, I don't know that I'm "totally off base". If I had been right that the "knock correction" value indicated the amount of timing the ECU was pulling in order to correct a knock condition, then I'd say I was pretty on the money. Don't blame me because ECUExplorer/Subaru decided to give the value a poor name.

[cody]

Originally Posted by sperry

Well, I don't know that I'm "totally off base". If I had been right that the "knock correction" value indicated the amount of timing the ECU was pulling in order to correct a knock condition, then I'd say I was pretty on the money. Don't blame me because ECUExplorer/Subaru decided to give the value a poor name.

Because "not knock correction" just roles right off the tongue.

[cody]

The IAM is at 16 and almost no knock correction. I imagine this is promising?

[cody]

So I turned down the boost a touch. 17.5 is a bit high.

[knucklesplitter]

Originally Posted by cody

The IAM is at 16 and almost no knock correction. I imagine this is promising?

That little bit of correction down low is seen often. It's usually from starting a pull below 2000rpm, but in real world driving you don't usually (and shouldn't be) lugging the car hard down below 2000rpm.

[cody]

Thanks, and I'm very anal about not lugging the car at low RPM's except when doing pulls.

[cody]

I imagine it's still a good idea to stop by EQTuning and have Ed data log me at sea level again (including wideband data)? Or can we assume that a simple Inlet Pipe upgrade isn't going to cause me an issue, even at sea level?

[sperry]

Originally Posted by cody

I imagine it's still a good idea to stop by EQTuning and have Ed data log me at sea level again (including wideband data)? Or can we assume that a simple Inlet Pipe upgrade isn't going to cause me an issue, even at sea level?

I'm with Matt... it seems like there's no appreciable difference during open loop due to the inlet, and under closed loop the ECU should easily compensate for any small change in airflow.

[knucklesplitter]

Originally Posted by cody

I imagine it's still a good idea to stop by EQTuning and have Ed data log me at sea level again (including wideband data)? Or can we assume that a simple Inlet Pipe upgrade isn't going to cause me an issue, even at sea level?

I think it's fine. But ask Ed and see what he thinks.

[knucklesplitter]

Originally Posted by sperry

Well, I don't know that I'm "totally off base". If I had been right that the "knock correction" value indicated the amount of timing the ECU was pulling in order to correct a knock condition, then I'd say I was pretty on the money. Don't blame me because ECUExplorer/Subaru decided to give the value a poor name.

The nomenclature was probably made up by Ecutek way back when, and then carried over to all the open source programs.

On later models it's called "knock correction advance additive". It's the advance that is added to the base timing when there is no knock correction. I just call it the the "additive advance" or "additive timing".

[renosubby]

Looks ok, but you could begetting some FBKC on a couple pulls... try logging that because Fine learning knock correction is only correction displayed by the ECU in the learning view. Reason I say this is because it was showing up to 8 in knock correction but it later goes down to 6 in the same area. so it could be knocking some.
Ed usually keeps around the same # so the tune is more consistent.

Cory

[cody]

What software should I use to log FBKC?

[renosubby]

Originally Posted by cody

What software should I use to log FBKC?

I can log you of you want me to. Just lmk

[cody]

Sure, do you have a wideband we could use?

[renosubby]

of course... fri-mon im open

[cody]

Originally Posted by renosubby

of course... fri-mon im open

YGPM

[A1337STI]

New inlet pipe for the pax win?? (or just cause your old one had a tear in it)

[cody]

Haha. I think I am seeing a touch faster spool. Hard to say with all the heat. I should be able to tell when I autox it though.