[JJ18Sniper]

Originally Posted by nhluhr

it's because your ECU senses the big FMIC and says "okay here we go, a big ass turbo is about to spool" and then it only gets the VF39. You need a SZ49, dawg.

okay sorry, i know that doesn't help.

I think a 3" 20G 8CM is moreso up his alley

[crystal_Imprezav]

Hey wolf, probably a dumb remark but are you sure you have -70 or -80% in your FIBTE in enginuity? Just double checking everything that it could be.

[WolfPlayer]

Originally Posted by crystal_Imprezav

Hey wolf, probably a dumb remark but are you sure you have -70 or -80% in your FIBTE in enginuity? Just double checking everything that it could be.

Absolutely 100% positive. In fact, I created 2 .xml files. On of those files kept the stock .XML settings and I had -80% in there. The other XML file used Cobb's way of doing it and I had (25) in that file (as opposed to the stock 128 value that is equivalent to 0% change).

I also took my original Cobb ST image (which Christian had modified the FIBET to -30%) for me and when I opened it up it said -30% across the board).

Can you tell that I am checking every damn thing possible? I just about give up - lol. I've tripple checked like everything.

The only other doubt that I have right now is that maybe the very quick filling of the pipes and intercooler by the VF39 causes a surge of air that could cause a boost leak (that would not be evident with me pressurizing the pipes and intercooler with the pressure tester -- but the MAF volts vs. Boost Pressure in the logs don't seem to bear that out).

t

[twiSTies]

You should just put an LS7 in there. That should give you enough airflow for that intercooler.

[crystal_Imprezav]

Ditch the APS and go SSAC ftw! haha, I have no more ideas at this point. I would say it has something to mostly do with the injectors being that is the only major difference in out setups. Maybe if I can poke around in your map either in enginuity or st, I may be able to spot something, email it to me if you want.

[crystal_Imprezav]

Oh yea, one other thing that I did today was move the BPV closer to the IC piping. As the stock STi is programmed for the BPV to be directly located on the IC itself, the metered air the ecu is probably used to seeing is probably drastically reduced due to the extra hose used to remount the BPV with the FMIC. I had a 90* elbow on mine and today switched it to just a very short hose running straight/perpendicular to the BPV and it has helped a noticable amount with the afr's in between shifts. I was looking at your APS setup and the hose is EXTRA long that connects the IC to the BPV. Try using some short tubing you may have lying around (I used extra radiator hose I had from the 90* elbow I bought) and see if that helps at all. Its a 1-1/4" dia hose. It also requires a minor adjustment of the hose that reconnects back to the inlet to make sure its not kinked up and maybe some longer vacuum hoses if yours are too short like mine were.

[SCOOBYTRAP]

Originally Posted by crystal_Imprezav

Oh yea, one other thing that I did today was move the BPV closer to the IC piping. As the stock STi is programmed for the BPV to be directly located on the IC itself, the metered air the ecu is probably used to seeing is probably drastically reduced due to the extra hose used to remount the BPV with the FMIC. I had a 90* elbow on mine and today switched it to just a very short hose running straight/perpendicular to the BPV and it has helped a noticable amount with the afr's in between shifts. I was looking at your APS setup and the hose is EXTRA long that connects the IC to the BPV. Try using some short tubing you may have lying around (I used extra radiator hose I had from the 90* elbow I bought) and see if that helps at all. Its a 1-1/4" dia hose. It also requires a minor adjustment of the hose that reconnects back to the inlet to make sure its not kinked up and maybe some longer vacuum hoses if yours are too short like mine were.

^^^This sounds promising!

[WolfPlayer]

Originally Posted by crystal_Imprezav

Oh yea, one other thing that I did today was move the BPV closer to the IC piping

Thanks a bunch for giving me something else to try. I will definitely try that. My next test is ready to go. I changed the FIBET to 0 (-100%). I verified this value with my different XML files. I'm not expecting it to work but I can't just sit here idling (lmao).

It's cool that you gave me something new to try. Only issue is, I get the low AFRs on spoolup (no shifting) too.

I also noted that this seems to be happening at not only WOT but also on part throttle accelerations. Jumping to 8psi and accelerating I see the AFRs go lean (need more tip-in) but then immediately rich (to 11:1) before settling back to something reasonable. This also seems to point to NOT having a boost leak. Air is just getting there before the fuel.

I can send my hex over to you since you already have ST.

t

[Guest_1u53r]

Maybe there is a table that we don't know about that compensates fuel for WGDC correction... meaning, perhaps the WGDC you're using isn't appropriate for the Target Boost you're using and the resultant corrections are causing unintended enrichment.

[WolfPlayer]

Originally Posted by nhluhr

Maybe there is a table that we don't know about that compensates fuel for WGDC correction... meaning, perhaps the WGDC you're using isn't appropriate for the Target Boost you're using and the resultant corrections are causing unintended enrichment.

Linking fuel to WGDC is an interesting idea. But, I don't think that applies for me (JMHO). I'm never in a situation where the ECU is trying to pull out any significant WGDC that was just added in. If you take a look at the log I posted on page 3, you'll see the car goes crazy rich even while the ECU is still trying to reach the target boost (turbo dynamics are positive here).

I'm still betting on a enrichment vs change in load (or MAF or MAP). It's in too many OEM calibrations to not be in this one.

t

[WolfPlayer]

This thread is getting so long that I don't think people will even read it this far - lol.

With a FIBET of 0 (-100%), and a stock tip-in (which is ludicrous given that the JECS need more tip-in down low), the car is 100% the same. ~9.5:1 AFRs on spool.

I have now expanded my search for this issue to part throttle as well. Guess what? I notice the same damn thing at part throttle . Specifically, I notice that when I am cruising on the highway, and I give the car a little gas, the AFRs first go slightly lean (not enough tip-in) and then a rich spike of about 1.0 AFR, and then the AFRs return to my fuel map values - lol. This is all open-loop with no fuel corrections (0.39/0.39).

The part throttle data reinforces that I don't have a boost leak (which we already know because I pressure tested the system with 25psi). If it isn't leaking at 25psi it sure isn't leaking at 5psi (rofl).

I don't have anything else left to try really. I can do the BPV relocation suggested by crystal. However, this is happing on spool everywhere (not just between shifts).

The next thing is to remove the FMIC top pipes, install the TMIC, and ensure that everything works as it is suppose to be working. Yay. If it does, then I might be looking at taking the topside cold charge pipe to have a MAF adapter welded on to go Blow-Through (much to the chagrin of RainMaker who is hating his blow-through.

* I thank everyone who has been working through this with me. Christian (Cobb), Daemon, Crystal, overboost & Ed (Eq), SS (SOG), Nick, etc.

t

[Siegel Racing]

It is "MAF lag." You are absolutely correct that there is a HUGE difference in time from when the air comes through the maf to when it goes into the motor.

Stock expects:

1) Floor the throttle
2) turbo spools pulling air past the maf
3) air travels 2 feet through a 60-65mm inlet to turbo
4) air travels 18" through small Y-Pipe
5) air travels one foot through TMIC
6) BAM past the TB and into the motor

With a FMIC it changes to:
4) travels 5 feet through big piping to fmic core (tmic has alread hit TB)
5) travels through 4 feet of core (car has been spraying fuel since 4)
6) travels another 5 feet of piping back to TB

Much of what has been discussed here has been discussed among tuners for years. You are certainly working toward "best case without blow-through." The nice thing with meth-injection is that it's boost based, so you don't use the maf to add a bunch of fuel pre-maturely... the meth comes on with boost in the intake (post TB)

Most tuners that have done blow-through don't really continue suggesting it. It's not "bad" and it can certainly be made to work, but it's pretty finicky...

SS

[daemon]

Seems like there has to be an ECU setting to modify this latency. I can't imagine the ECU programmers hard coding a value like that. I guess we just need to find it.

Heck, even if they did hard code it, it still has a memory location and can be altered.

[Siegel Racing]

There should be a "maf delay" line of code in there somewhere.

SS

[crystal_Imprezav]

Tim, you can send your hex and st map to me and I will do some testing with it, the only problem is I will be in Texas w/o my car until the 5th. This will however give me plenty of time to thourghly go over it in both enginuity and st while I am at the airport and on the plane and the next week while I will most likely be bored out of my mind during interviews. I would suggest the BPV shortening. It is at least worth a shot IMO along with some adjustments to the very lower part of the MAF table (.94-1.17v). This has drastically helped the in between shifts afr's as I am now back to around a 12:1 afr going from a 10:1 before BPV move and 11:1 after move. Things are almost back to stock like feeling minus the need to shift extra quickly and have more throttle between shifts almost like flat foot shifting which I now have my TDC table setup to do.

Another thing I discovered yesterday that I have been painfully staking over was why my defi boost gauge was soooo damn slow reacting and the troubles I was having dialing in my meth injection. F*ing Subaru installed some pill like restrictor on the vacuum line off the BPV T when they installed the factory boost gauge that you can barely see the split in the line with everything connected. I actually came across this accidently as the vacuum line came off and apart while I was moving the BPV and I had to extend the vacuum line a couple of inches. I was like WTF is this!?! It looks like a freaking vacuum line connector (brass) with one end a big opening and the other end a pin sized opening. I have never heard or seen anything like this the dozen of times I have been under the hood of a subaru. I removed it, extended the vacuum line a whalaa the boost gauge was like lightening fast like everyone else's and I am sure now that when I get the meth back in, it will properly kick on when it is suppose to a getting the fuel finally dialed back in will be a piece of cake.

I am sure that the extra piping in the APS unit, along with larger pipes (i think) and a much larger core is the reason you are seeing a larger afr dip than I am. The APS unit is much too large for a stock turbo IMO and should be used with at least a 20G min. IMHO. I say sell the APS IC spend the $300 on the SSAC IC and used the extra $800 or so on something better i.e. the new turbo fund, ALCOHOL INJECTION, inlet, headers, etc. I am sure a table for MAF lag/delay or what not will come out soon in enginuity since they are working on xml's that will distinguish when the ecu will use what table (i.e. DA A,B,C) and way's to calibrate those tables such as the ecu will only use DA A like cobb already has control over. Another thought may be, that maybe adding some extra delay into the closed loop delay table will able you to delay when the ecu goes into open loop, therefore giving the car time to get the air to the TB before it switches into openloop giving you a chance to some better afr's. Maybe at this point we can play around with the first two values which are left at 0 while in stock form the third value is 313. Adding some delay to all three may discover something to help our cause. The more ideas the better, that is if your like me and don't care how many basemap flashes you flash.

J