Full tune of 68HTA, KSTech 73 MAF, Racer X FMIC and DW750s

[LittleBlueGT]

RR allows for the PLX SM-AFR...

 

Really. I have the 0.5.2 version and it only says innovative and AEM.

 

I think the model I have is the M300 or something like that. It was specifically designed for ST.

 

If you know how to get it work with RR, that would be awesome!

[LittleBlueGT]

Right now I am at 806 for the scalar (COBB 3361) and my latencies are 3% higher then OEM. This is my result:

 

http://i248.photobucket.com/albums/gg194/littlebluegt/LearningView_SS_9-4-200941039PM.jpg

 

So I am pretty close. I need to tweak the latencies a bit more.

 

What have others gotten with the DW850s?

 

If it seems like I am taking my time with this, I am. I don't have everything yet to do my turbo swap, so I will slowly tune my inj and intake in prep for it, but I refuse to drive just for the sake of tuning. I will only tune if it is on my way.

[NSFW]

Really. I have the 0.5.2 version and it only says innovative and AEM.

 

I think the model I have is the M300 or something like that. It was specifically designed for ST.

 

If you know how to get it work with RR, that would be awesome!

 

There's a pre-release version that supports the SM-AFR, there just hasn't been an official release since that code was added. I wrote the basis for SM-AFR stuff - Kascade ported my little C# example to java. I also had a look at the 300/500 protocol, but unfortunately it's not compatible and I don't have one to test with. Supporting a wideband that I can't test with would take while, but if you are interested in testing buggy software once or twice a week, I could take a shot at it. Maybe we could have it working by Christmas.

 

I'm guessing that an other .1 millisecond of latency will bring your A trim pretty close, but also raise your B trim to about 5%. (I've been futzing with mine lately too.)

[LittleBlueGT]

I will keep it in mind, pm the details of what I would have to do.

 

I guess that some more latency will bring A in line, but I also think it will lower B closer, not raise it.

[LittleBlueGT]

My injectors are almost done. I had to runa fair bit more latency then stock. Even though DW says they are like stock. My inj scalar is gonna end up being close to 806 or so. I think it is odd that Subaru has the exact same injectors in the 05 LGT as the 05 STI, but the scalar is 520 (LGT) vs 550 (STI). So the whole things is likely wacked a bit. It doesn't really matter that much, but it is nice to have more exact numbers, then you are more likely to hit your AFR targets from the pri fuel table.

 

In the future I may try putting in 867 for the scalar, stock latency, then tune the MAF that way, see it makes much difference (especially in the high rpm/low load areas).

[LittleBlueGT]

To keep you guys occupied I decided to reveal a whole bunch of tuning secrets.

 

Here are some screen shots from RR, I will try and explain the changes I made, and why I did so: (From top left to bottom right)

 

http://i248.photobucket.com/albums/gg194/littlebluegt/RRscreenshots1.jpg

TD Integral Cumulative Range: Just so people know you can change the td integral (COBB speak = coarse gain high/low): Right now I have the positive set to max out at 10. When I was running IWG I had it much higher, as I had some creative td values to prevent boost spiking.

 

MAF Limit: OEM is set to 300 g/s, if you are gonna hit higher then that (especially on those cold nights) set it higher!

 

Engine Load Limit: Same thing as MAF, set it high enough, with room to spare!

Pri OL Fueling (Failsafe): This is on my alky tune, so my failsafe is about where my pump tune is at, just a hair lower (richer). More on the switch later.

Boost Control disable (IAM): I always tune for and run my DAM/IAM at 1.000, therefore, if it falls, even the slightest due to knock, I have it set to disable boost right away!

 

Pri OL Fuel Map Switch (IAM): I also have this set very high, again I always run at 1.000, but if it falls (like due to an alky system failure) my failsafe map comes on right away.

Throttle tip-in & comp (boost error): I copied these tables from the 05 STI (tip-in is now much smaller due to DW850s), they are set-up differently in the 05 LGT ROM. I can't honestly remember the why, but I figured since my TMIC and turbo size were now closer to an OEM STI that these tables might work better.

[LittleBlueGT]

Next:

 

http://i248.photobucket.com/albums/gg194/littlebluegt/RRscreenshots2.jpg

 

OL/CL delays: are all zero'd, many tuners do this, much easier IMO. Some tuners fell if the car is in many different environments (and fuels) that a small delay should be left.

Timing comp Intake Temp: I changed this table a bit, smoothed it out a bit (IIRC there is inexplicable a weird bump at 68F on the OEM tune). Since this is my alky tune I do not pull near as much timing when IAT skyrocket. If you don't run alky, and you run a TMIC, you may need this to help overcome TMIC heatsoak.

Timing comp....B: this table is kinda weird from factory, it adds timing up until about 3300 rpms or so (others know the exact details). Many zero it out to make things more consistent

 

Timing comp per cyl: this is in beta stage still (not sure which cyl corresponds to C, D or E) I felt that D was cyl #4, so instead of adding 2.11 degrees, I only add a little, to try and put any knock on the cylinder where the knock sensor is. This is still open for debate, but it is my theory. All other tables (A,B,C) are zero'd

[LittleBlueGT]

http://i248.photobucket.com/albums/gg194/littlebluegt/RRscreenshots3.jpg

 

FBKC range rpm: change it to where you will run the motor

FBKC retard value: I figure that when pushing more air through an engine, you run less timing, therefore one should pull less timing if it knocks.

 

FLKC range rpm: change it where you will run the motor

 

FLKC rows and columns: again, change it to where you will run the motor, make it applicable, especially if running much higher load/rpms

 

FLKC range load: THIS IS A VERY IMPORTANT ONE!!!!!!!!!!!!! I have seen countless tunes from COBB (Christian has responded to my e-mails, but he doesn't seem to get it:spin:) COBB has the max load value for FLKC set too low for almost all of their OTS tunes IMO. If you can hit higher load then this table goes up to, then you have very little protection there. ie: tuning an 05 LGT the OEM value is 2.2, COBB stage 2 also uses 2.2! You can easily hit well over 3.0 load on a COBB stage2 tune on a LGT. What this means is that if you knock above 2.2 load, the only form of protection your motor has against knock is FBKC. But FBKC is never learned, it does not affect IAM/DAM, so unless you log the actual knock event you will never know it is happening. You could knock every single time you go above 2.2 load, and the ECU will respond with FBKC, but it will never show up in FLKC, and thus will never affect IAM! Bad bad bad!

 

FLKC retard & advance: copied from NSFW, reducing timing by 1.01 and then only adding it back by +.25 will enable one to see how many times it knocked. ie: if it knocked twice, but then had 4 consecutive adds, it would be -1.02.

 

Rough correction tables: same things as other ranges (FLKC, FBKC) I raised the min, as I didn't want low load knock to lower IAM

Advance Multiplier (Initial): I set it to 1.000. I think it is easier, that is what I tune for. More then one school of thought here.

 

*** a note about my settings*** I have a clutch/pressure plate combo that makes my knock sensor go nuts at lower loads. I thus have many settings set a bit higher (like low range load/rpm) then OEM to stop this from affecting driveability. I have checked many times to ensure that this knock is not real, but from the drivetrain.

[LittleBlueGT]

I got more, but I am tired.

 

As always, if you are stupid enough to use my settings w/o much knowledge, or if it causes your car to blow up (causing a semi to swerve hitting a limo, that has the president in it, thus causing WWIII) it is your own fault!

[ClimberDHexMods]

Great posts, learned a lot!

 

My $.02: get a P&L oil feed line and install it at the same time as the hta68. That way you won't need to worry. FP tells me they are also making a line soon that will allow more flow to the journal bearing chra. If it saves your turbo, why the heck not

[LittleBlueGT]

I am thinking about it. The guy at FP said he would take out the banjo bolt filter and replace the bolt w/ a less restrictive one (which is included in the kit) if it were his car.

 

I anticipate showing pics of both OEM and FP banjo bolt.

[ClimberDHexMods]

That filter bolt is not the restriction, in any circumstance. The pinhole sized tube entry leading to the turbo is the restriction once the bolts are taken care of. There are threads with the picture of it somewhere. Robert and I spoke about this for a long time. Basically he said the danger in keeping a restriction in there is that multiple back-to-back runs could deprive it. Drilling out that pinhole was another option, but getting the line off/on requires stripping the top of the passenger head bare. But then again, has anyone heard of this turbo dying yet? I'm more worried with my 20g, a bigger chra...

[NSFW]

I guess that some more latency will bring A in line, but I also think it will lower B closer, not raise it.

 

Yep, you're right about B, my bad.

[dschultz]

Great write up!

[Infamous1]

I think it is odd that Subaru has the exact same injectors in the 05 LGT as the 05 STI, but the scalar is 520 (LGT) vs 550 (STI). So the whole things is likely wacked a bit.

I think I may have mention this to you before, but the 04-06 STi scaler is actually lower than the 05-06 LGT, (503.93).

[LittleBlueGT]

I think I may have mention this to you before, but the 04-06 STi scaler is actually lower than the 05-06 LGT, (503.93).

 

 

You are correct. I was looking at the EDM market STI, which shows 550.67. I wonder if the injectors are the same?

[LittleBlueGT]

You are correct. I was looking at the EDM market STI, which shows 550.67. I wonder if the injectors are the same?

 

 

Interesting, 09WRX uses same scalar as 06 EDM STI, maybe the 06 EDM STI uses top-feeds?

[LittleBlueGT]

last one:

 

http://i248.photobucket.com/albums/gg194/littlebluegt/RRscreenshots4.jpg

 

Requested Torque: look at last line, that is a soft rev-limiter.

 

Rev Limit Fuel Resume (Boost): OEM is 9ish psi, I set it to 17. When I hit the rev limiter, it will resume fueloing when it gets below 17 psi. OEm settings make rev-limiter feel like you broke somthing.

 

And last, but definitely not least:

 

Speed Limiting: I am not 100% sure on the aerodynamic stability of my car above 186 mph, so I decided to keep things on the safe side. I may be an Air Traffic Controller, but I am no pilot!

[Legend]

Good thread, LBGT. You've reminded me that I've got a couple things to do next time I resume tweaking my tune... namely the FLKC load range and the "soft rev-limiter".

[LittleBlueGT]

namely the FLKC load range

 

The clue I always see that the load range for FLKC isn't high enough is this:

 

-a log that shows steady load (turbo already spooled) but the ECU activates FBKC.

 

For me 99% of the time, if I get a knock event in a fairly steady load situation (turbo spooled, not right after a shift, not in a transient situation) it will show up in FLKC. One event, it will be there.

 

I only get FBKC in transient situations.

[Guest]

Interesting, 09WRX uses same scalar as 06 EDM STI, maybe the 06 EDM STI uses top-feeds?

 

I am not positive but I think JDM GD STI was always top feed as well. Interestingly JDM BL/BP LGT has remained side feed throughout all years.

[LittleBlueGT]

About the injector scaling. I think I will use the numbers that DW gave me.

 

I recall (after scaling my MAF) noting that the AFR at high rpm/low load was not matching up with my pri fuel table. I don't recall which way it was off. So I will try using the flow numbers from DW, then scale the MAF (got my K&N all in today). I will then see how close thew high rpm/low load area is. And maybe I will tweak my latencies from there.

 

Or maybe I am on a useless pursuit of perfection that is unattainable.

[LittleBlueGT]

Actually did some real logging today and tuning of latencies. My conclusion, before telling all the facts (this is the internet) is that the stock ECU does not correctly tune injectors.

 

On my previous tune I had stock injectors, and never did touch the settings. I just scaled the MAF, and went on my merry way. I did notice however that the high rpm/low load AFRs were substantially off from the primary fuel table.

 

Today I did some very steady logging, I did some 3rd gear pulls and kept MAFv very steady, but allowed rpm to climb. Simplifying things a bit, I consistently got about 7 to 9% errors when comparing the same MAFv, but at drastically different rpms. ie: 2.5 volts show rich at low rpm (2300), but lean at high rpm (5000). 2.8 volts, same thing, etc........ This was despite the fact that my latencies were a bit higher then OEM tune.

 

This is what I ended up with (my errors were within about 1-2%):

 

 

Stock latencies:

 

http://i248.photobucket.com/albums/gg194/littlebluegt/Enginuitylatency.jpg

 

My new latencies:

 

http://i248.photobucket.com/albums/gg194/littlebluegt/DW867latencies.jpg

 

 

 

Don't look too closely at the low voltage part of it, I took an educated guess there. At 14 volts it is about a 19% difference from stock! To me that seems like a lot, but I do recall the stock one being out as well, how much I don't know, but I am certain it was in the same direction (stock latencies too low).

 

My testing was done at +/- 4C IAT, all within about 1 hour of each other. I looked at all points in the log (after sorting via MAFv) and did not even log anything that was not steady state throttle (ruling out tip-in). I changed Min Active Pri OL enrichment to 14.65, and ran all of my targets at about 14.25 or less to keep things below 15:1 AFR (stays in OL the whole time, no learning).

 

Since then I have plugged in 867 in the scalar, and I will now scale my MAF. I hope this will result in a primary fuel table that (+/- 2%) equals my targets.

[John M]

I'll wait till you get them dialed in and then borrow your settings

 

I was waiting to get my intake set up on stock fuel and then put in the DW850s to tune them.

[ClimberDHexMods]

I keep coming back to this thread, learned a lot from it, thank you for writing it!!!