I've been discussing my tuning issues with a couple of you Enginuity gurus, via email, concerning my issue I'm having enormous trouble resolving.. I thought I'd post this to the forum, in case someone has dealt with this scenerio in the past, or in case someone else runs into this issue..

Here's the parts list:

EJ25 block/crank
Arias 9.5:1 pistons
Eagle H-Beam cromoly rods
ARP head-studs and rod-bolts
Cometic Head Gaskets
TGV deletes
Modified stock Injectors (~830cc, decapped)
Precision GT32/55 turbo (48lb/min)
Rotated uppipe
Tial 46mm Ext. Wastegate 1.1bar spring
Tial 50mm BOV
Precision FMIC
2.5” hot-side FMIC piping
3” cold-side FMIC piping
3” intake (MAF in 3” pipe)
Aeromotive fuel rails
3/8” fuel lines
STI FPR
Stock cams
3 angle valves, blended bowls
AR Fab modded oil pump (190*F, 70psi idle, 100psi @ off-idle)
Walbro 255lph, 42psi idle


When cold, it starts right up, idles fine. As it warms up and idle starts to drop, if begins to get rougher and rougher.. It seems to drive "ok" through town, with a slight miss here/there.. BUT, when accelerating, and the transition from vacuum to positive pressure takes place, the car goes full-RICH, instantly.. Logs of both the WBO2 and the AF Sensor #1 confirm this. I'll add that when this occurs, the car stumbles a good bit, obviously due to the huge amount of fuel, I'd assume.

Cruising around town, steady 30-40mph, it reads 14.8-15.3 AFRs..

Ive rescaled my modded stock injectors from 835 to 725, and multiplied the tip-in throttle enrichment table by .58 (420/725). It acted the same, when scaled to 835.
Ive got no idea what I need to do to correct this issue..
:sadbanana:Any ideas?

I’ve got a log from yesterday. In the attached snippet of the log, you can see an area where the AFRs go to 11 or richer.. I’m not really flogging the car, but more of just like 2nd gear, ½ throttle boosted acceleration.
http://img217.imageshack.us/img217/733/logshotyb5.jpg (http://imageshack.us)


the "wbo2" column is calculated from 0-5V input to ecu 'fuel level (V)' from my PLX wbo2 0-5V output, converted to AFR by =SUM(H2*2+9) equation. You can see from the log, that it's about a point or so, off of the AF Sensor #1 AFR reading.. Although, it is about 4' downstream of Sensor#1.

Well, at first glance I think you have your injectors scaled too small as the A/F correction is very negative, denoting too much fuel.

I would spend more time getting the injectors properly sized and the latencies correct. Have you adjusted the latencies yet?

Just a quick guess but your injectors should end up around 750-780 and the latencies need to be increased 30-40% or so. (I am assuming that these are modded WRX injectors as I seem to remember that you have a WRX). I tuned a set of modded injectors on an 03 WRX a while ago and they did take some time to get right.

You should also log your A/F Learning A-D over time. These all should be around +/-5% long term. If not then you need to tune the injectors more.

Also, how did you tune the new MAF pipe? Are you trying to tune both the injectors and the MAF at the same time? If so, it will be much more difficult.

PM me if you need more info or want to chat.

Thanks for the posting..
Last night, I rescaled the injectors to 750, and that was the only change.. Here's a screenshot of the log snippet, where boost transition occurred.. In addition, the car will not idle on it's own, with the injectors scaled from 725 to the new 750.

They are, indeed, modded stocker WRX injectors.

I haven't tuned the MAF pipe yet, either. Unfortunately, I am trying to tune the two, together, which as you said, has proved to be quite difficult: especially given the fact that I, too, am on a tuning learning curve with this! lol

http://img217.imageshack.us/img217/3913/logsnapshot2nh3.png (http://imageshack.us)

Tuning larger injectors and a big maf at the same time is a pain. Just find the scalar/latency values that make your idle good and scale the maf from there. Please note that changing your MAF scaling will cause your engine load to move around so watch out for that. You don't want to move into a low load column with a lot of timing in it at high boost. You are running very little timing at 15psi and very rich too. I run 11.1:1 and 16-17* at 4500rpm and 22psi.

the "wbo2" column is calculated from 0-5V input to ecu 'fuel level (V)' from my PLX wbo2 0-5V output, converted to AFR by =SUM(H2*2+9) equation. You can see from the log, that it's about a point or so, off of the AF Sensor #1 AFR reading.. Although, it is about 4' downstream of Sensor#1.According to http://www.plxdevices.com/M-Series-Controllers/WidebandOutputGraph.jpg, your conversion from voltage to AFR is incorrect and produces AFRs that are one point richer than the actual AFRs

here's a snapshot of the ign table... looking at log, where my AFRs go from 11s to 9s, and matching that area to the IGN table, I see that timing is greatly reduced there..

http://img227.imageshack.us/img227/9792/750ccigntabletw5.jpg (http://imageshack.us)

;20582303']According to http://www.plxdevices.com/M-Series-Controllers/WidebandOutputGraph.jpg, your conversion from voltage to AFR is incorrect and produces AFRs that are one point richer than the actual AFRs
thanks, jon. I initially started out using "10" (I think about a year ago, you helped me with the conversion, initially, on enginuity forum.. :) ), then somewhere down the line, I got messed up and used "9". I'll redo the conversion. It's still extremely rich ;)

Watch out having too little timing. It'll make your egt's go up.

yeah, i noticed the EGTs high, last night.. well, relatively speaking.. sitting in the driveway, holding throttle to 2500, the EGTs went to 1350. probe is 4" from the head, on #4 cyl.
Idle sits around 8-900F, which is higher than I'm used to seeing..

post #3 is updates screenshot with wbo2 calcs corrected to "10", nstead of the incorrect "9"..

here's a snapshot of the ign table... looking at log, where my AFRs go from 11s to 9s, and matching that area to the IGN table, I see that timing is greatly reduced there..

http://img227.imageshack.us/img227/9792/750ccigntabletw5.jpg (http://imageshack.us)

weirdest looking base timing I have ever seen - what is the advance table like - you must have some huge values in there?

looks to me like he's using 10* of advance throughout the entire timing map.

Let me start by saying I'm a noob when it comes to this stuff...perhaps your throttle tip in needs tweeking. in both cases, you are going almost 100+% increase in throttle quickly and then the car goes rich. log the tip in % and "last calcualted enrichment"(or something like that). you may want to decrease the tip in and see if that helps. However, I would imagine that as you start to get your injectors and MAF set correctly, you may also need to keep adjusting the tip in....so like it was said before(maybe another thread), put your factory air box on first...get it settled, and then switch.

looks to me like he's using 10* of advance throughout the entire timing map.
the entire Timing Advance (Maximum) table is 9.4

Let me start by saying I'm a noob when it comes to this stuff...perhaps your throttle tip in needs tweeking. in both cases, you are going almost 100+% increase in throttle quickly and then the car goes rich. log the tip in % and "last calcualted enrichment"(or something like that). you may want to decrease the tip in and see if that helps. However, I would imagine that as you start to get your injectors and MAF set correctly, you may also need to keep adjusting the tip in....so like it was said before(maybe another thread), put your factory air box on first...get it settled, and then switch.
The tip-in has been multiplied by .56, already, but that doesn't make it right, I know :)

As for the factory airbox:

http://img184.imageshack.us/img184/9048/dec24cib6.jpg

Not really an option :sadbanana:

As for the factory airbox:
Not really an option :sadbanana:

nice setup, since the box is not an option...good luck.

as for the tip in table, start pulling it back :)

I may be wrong here, and just my opinion, but hopefully, your MAF behavior will be consistent. if so, I would leave it alone and just tune around it. think about it, load is based on RPM and MAF...so as long as both are consistent(right or wrong) the map should reflect it, perhaps in "lumpy" timing and a very lumpy fuel table....but that will then show where the focus needs to be.

I've heard of this happening on FMIC cars where the stock MAF placement is retained, but I would think that behavior would not be found on a blow-through setup. I'm eagerly awaiting some expert comments on this one, as I'm in the same boat, albeit with a stock MAF location.

Ben,
My maf is geograhpically in the OE location, give or take an inch.. It's not blow-through, either. I was going to go blow-thru, but heard it's a bunch of headache to tune.. look at me now, go figure! lol

If it's continously rich during open loop fueling then it's a Fuel injector scale issue. If it's just a spike that returns to normal then it's the problem a bunch of us have been dealing with.

Hmm, I don't have my tuning laptop with me, but I recall a map in street tuner that was like tip in enrichment, except for changes in manifold pressure as opposed to changes in tps. Cobb calls it fuel injector boost error trim, i'm not sure what enginuity calls this table, but its worth looking into.

Found this little gem, which should apply to you with the stock location maf. More here: http://www.cobbforums.com/forums/images/AccessTUNERSoftwareCalibrationTuningGuidesSubarusv 111.pdf
IF YOU ARE RUNNING WITH A FRONT MOUNT INTERCOOLER, WE SUGGEST YOU SET ALL OF THESE VALUES TO ZERO (see below
table). This will help prevent over-fueling the engine due to the time delays associated with having the MAF sensor located further away from the throttle
body. With a front-mount intercooler installed, you have dramatically increased the length of intercooler piping. The MAF sensor meters the additional air
coming into the engine before the air actually reaches the engine. This can create a temporary over-fueling of the engine until the increased air charge
reaches the engine.

*edit: enginuity calls it Throttle Enrichment Compensation (Boost Error).

Good luck

thanks for the interesting link and info! :)

Cobb's description of FIBET is incorrect.

It is not based upon rate of change of boost.

It is a multiplier against tip-in based upon boost-error. That's what merchgod said based upon his review of the rom logic. I believe he has it in the defs "help" description as well.

Much better logs posted here than the ones you sent me. The fuel table is doing exactly what its told to do. You need to raise the injector scalar to 785-800 & increase the latency to get the idle under control. You are only ~1.0 AFR off from what the fuel table is calling for @ >2.0 load. The fuel log vs. Load log is pretty smooth considering the setup isn't dialed in.

Much better logs posted here than the ones you sent me. The fuel table is doing exactly what its told to do. You need to raise the injector scalar to 785-800 & increase the latency to get the idle under control. You are only ~1.0 AFR off from what the fuel table is calling for @ >2.0 load. The fuel log vs. Load log is pretty smooth considering the setup isn't dialed in.
thanks, jeremy. Actually, post #3 log snippet is from one of them I sent you on Wednesday ( enginuitylog_20080102_175043.csv)

you've definitely been a big help with this. I wasn't doing any name-dropping, as I didn't know if you'd get bombarded with messages/emails from a bunch of people wanting help, and I didn't wanna be the root of the reasoning :)

Cobb's description of FIBET is incorrect.

It is not based upon rate of change of boost.

It is a multiplier against tip-in based upon boost-error. That's what merchgod said based upon his review of the rom logic. I believe he has it in the defs "help" description as well.

This doesn't come as a surprise, I honestly think cobb doesn't know its own ass from a hole in the ground. Heck, their name 'fuel injector boost error trim' itself points toward merchgod's definition.

thanks, jeremy.

no problem.....you are not as far off as you think.

#3 post log using AFR table for 'Target AFR'
http://www.wrxfanatics.com/uploads/monthly_01_2008/post-367-1199484654.gif

Depending on what the scalar was for that log, you need to raise the scalar ~10cc per every .1-.2 AFR off (you're roughly .4 off) from target then dial in the latencies & the MAF table for idle.

their accessecu tuning guide also has many interesting holes. for one, it has a provision to calculate rough mass air flow scaling. the excel computes the difference as a change in area (pi * r^2) but the formula in the description calls for a change in surface area (4 pi * r^2).

sheesh! how the hell you graph that? I've been sitting here for the past 1.5 hrs in Excel (2007 for Vista) trying to figure that out.. :furious:

Are you using Airboy's interpolation spreadsheet. It looks like it to me.

Are you using Airboy's interpolation spreadsheet. It looks like it to me.
I just tried searching nasioc for said airboy spreadsheet, but "search is disabled."

:diaf:

It is a must for tuning timing... at least for me.

You can also use it to tune open loop fueling.

http://www.ken-gilbert.com/wrx/enginuity/Log%20file%20Interpolation%20ver3p6L.zip

thanks!

+1 on the airboy spreadsheet, just select the log, click a few buttons and you get the graph testes posted and then some...

their accessecu tuning guide also has many interesting holes. for one, it has a provision to calculate rough mass air flow scaling. the excel computes the difference as a change in area (pi * r^2) but the formula in the description calls for a change in surface area (4 pi * r^2).

Yea, I posted about this in the thread they posted the guide in many weeks ago. Still no response.

Same goes with their definition of turbo dynamics.

Are you using Airboy's interpolation spreadsheet. It looks like it to me.

no, it only took ~20sec to create that graph but took 5 minutes or so to get the Map target accurate. I've never used Airboy's sheet. Its too cumbersome for me to use, at least one of the versions I used was, no offense.

no, it only took ~20sec to create that graph but took 5 minutes or so to get the Map target accurate. I've never used Airboy's sheet. Its too cumbersome for me to use, at least one of the versions I used was, no offense.
yeah, I tried using the spreadsheet from the link, last night. spreadsheet > me...

I couldn't get it to do squat, even remotely. PBKAC error, I'm sure. Even with reading the Info tab.

I do have to say though, Jeremy, looking at a log spreadsheet, and looking at your graph, the graph make totally more sense! LOL

worked a little more on the maps. It's getting better.. Still in the low 10 AFRs, but no more falling on it's face.

Just need a little more tweaking in the WOT settings, then start working on the idle and cruise..

now, if I could just figure out, solidly, Airboy's spreadsheet... is there a walkthrough, anywhere? LOL

Ive played with it and got verbal instruction from blkscooby, and read the first sheet like uberthousand times.. I still don't get it LOL

What version of excel are you using?

I had some issues with 2007 not allowing the macros to do their thing correctly.

2007...

Just a little update, the car is getting a touch better.. I really need to work on the idle and the "missing" or "stuttering" at cruise..
I've been staring at my MAF table, scared to change anything, as I am unsure how. blkscooby has explained it to me a couple times, but I just can't seem to get my head around it.. :(

Just a little update, the car is getting a touch better.. I really need to work on the idle and the "missing" or "stuttering" at cruise..
I've been staring at my MAF table, scared to change anything, as I am unsure how. blkscooby has explained it to me a couple times, but I just can't seem to get my head around it.. :(

Well, call me again if you need to :) Plus maybe I can help walk you through Airboy's spreadsheet that I sent you. Unless it is a Microsoft problem with Excel 07. Don't know. I use Excel XP.

ARG. still strugglin.. :furious:

hopefully this week, I can get it narrowed down some more. Gonna check for leaks in the charge-pipes later this week/weekend. I can "hear" what sounds like a slight leak on idle, and BARELY touch the throttle, and it instantly goes away, just haven't found it's source yet. "sounds" like on the drivers side, under the intake plenum, but it's so hard to tell exactly where it's coming from..

Make sure macros are enabled in excel, afaik they are disabled by default. Google should have some easy instructions to enable them.

Basics of the maf scaling: Where wideband afr is greater (leaner) than target afr, increase the maf value. Where wideband is lower (richer) than target, decrease. Basically, the maf table is a representation of how much fuel the ecu dumps for a given airflow in maf voltage.

If you get the spreadsheet working, you'll take logs of wideband, maf g/s, maf voltage, rpm, and load. Then import them, and import your fueling maps from enginuity, grab headers, get data, and interpolate afr. It'll spit out a nice graph of the afr error % vs maf voltage.

If you want to take small steps, just use half or quarter of the % error to adjust it, then do another pull.

Another option is to use Williaty's method without closed/open loop filter. IMO, it only takes a couple of tries using Airboy's spreadsheet to get it tuned.

Definitely check on the leaks, that'll throw everything maf related off, which is alot of stuff.

how far off are your maf AFRs from your wideband AFRs on the top end?

find somebody with a leak detector (either smoke or ultrasonic). It will expose leaks in no time.

I used an ultrasonic to find my torn fuel rail seal that only leaked vacuum/boost bad when the fuel pump was at full duty cycle (during startup and boost).

Make sure macros are enabled in excel, afaik they are disabled by default. Google should have some easy instructions to enable them.
They are..

Basics of the maf scaling: Where wideband afr is greater (leaner) than target afr, increase the maf value. Where wideband is lower (richer) than target, decrease. Basically, the maf table is a representation of how much fuel the ecu dumps for a given airflow in maf voltage.

If you get the spreadsheet working, you'll take logs of wideband, maf g/s, maf voltage, rpm, and load. Then import them, and import your fueling maps from enginuity, grab headers, get data, and interpolate afr. It'll spit out a nice graph of the afr error % vs maf voltage.

If you want to take small steps, just use half or quarter of the % error to adjust it, then do another pull.

Another option is to use Williaty's method without closed/open loop filter. IMO, it only takes a couple of tries using Airboy's spreadsheet to get it tuned.

Definitely check on the leaks, that'll throw everything maf related off, which is alot of stuff.
Thanks, I'll give it another go.

how far off are your maf AFRs from your wideband AFRs on the top end?
sorry, MAF AFRs?

find somebody with a leak detector (either smoke or ultrasonic). It will expose leaks in no time.

I used an ultrasonic to find my torn fuel rail seal that only leaked vacuum/boost bad when the fuel pump was at full duty cycle (during startup and boost).
I dunno anyone with either, unfortunately.. I'll be pressure-testing the inlet and hoping my hearing hasn't failed me beyond the point of finding any/all leaks.

thanks to all for the input! ;)

woops, I meant map AFRs, in other words, what does your fueling map call for, and what is coming out the tailpipe?


Regarding scaling your maf, I tuned an AEM intake on a WRX and found that the inner diameter was about 1mm smaller. Ended up using the formula pi * R^2 for both diameters and finding the percentage difference. I multiplied the MAF g/s values by that amount and it greatly smoothed out the large swings in correction, and got the trims within 5%. After that it was just a matter of traditional MAF scaling. Have you measured your intake pipe with a digital caliper or similar?

woops, I meant map AFRs, in other words, what does your fueling map call for, and what is coming out the tailpipe?

Ahhh! ;) I'll have to look again..


Regarding scaling your maf, I tuned an AEM intake on a WRX and found that the inner diameter was about 1mm smaller. Ended up using the formula pi * R^2 for both diameters and finding the percentage difference. I multiplied the MAF g/s values by that amount and it greatly smoothed out the large swings in correction, and got the trims within 5%. After that it was just a matter of traditional MAF scaling. Have you measured your intake pipe with a digital caliper or similar?

I haven't but will in the extreme near future. Prolly when I pressure-test the system.
I'll post any/all results

Keep us updated.... Feel free to call again if you need.

Bump, trying to fix this dip in afr after spoolup. I too run a FMIC, anyone try zeroing out boost error?

Bump, trying to fix this dip in afr after spoolup. I too run a FMIC, anyone try zeroing out boost error?
What's it doing, exactly, Infamous? where's the fatness?

Well on WOT pull from 2K it is from about 2600-2900 where AFR drops to 10.3-4ish with a 11.47 target. My Tipin enrichment has been increased however because I was getting a lean spike on intinial tip in to the throttle.

whatcha running for injectors?

Once past the dip, does the logs show it to be on AFR target?

Stock injectors, AFR is right on after the dip. Dip will occur at any point in the rpm range just after initial tip in. Tip in was initially setup for me with Mickeyd's help.

Welcome to the club :)

I've been dealing with this for a long long time (much longer than your initial post). I made one of the first posts about it on IWSTI awhile ago.

First and Foremost, do not set your boost error map to a rediculous value like -100%, etc. This will get you nowhere except to get rid of the throttle change acceleration enrichment (pump shot). That is bad ... and it is not what you need to do. The problem is not the pump shot. Your AFR is going rich on spoolup. That means that you can nail the throttle at say 1800 rpms and sit there and wait for spoolup. Pump shot is completely over before you get that rich spike. Yea - been there done that - lol. I tried tons of values in that table (and others) and you can't get rid of it. Read my posts on this page (by looking for my username: WolfPlayer) ...

http://www.romraider.com/forum/topic2627-60.html

t

. http://img217.imageshack.us/img217/733/logshotyb5.jpg (http://imageshack.us/)


I have the PSI3 ECU Data Monitor and my KC: Knock correction Degress (ignition offset). When I drive my Knock control Adv.goes9 like your is that right. Isn't suppose to be Zero knock at all time's I dont get. Iam' asking because I have to retune my car and I dont want to tow my car to Bufflao, I was just wondering should I drive it becasue the KC Adv. Degress goes up as I accerate up on the gas. But stay's 0.0 must the time as I'am just cruising. Just want to make on the Knock Control Adv. I make sure it's safe to drive it back to my tuner without towing that's it.

.
When I drive my Knock control Adv.goes9 like your is that right. Isn't suppose to be Zero knock at all time's I dont get. Iam' asking because I have to retune my car and I dont want to tow my car to Bufflao, I was just wondering should I drive it becasue the KC Adv. Degress goes up as I accerate up on the gas. But stay's 0.0 must the time as I'am just cruising. Just want to make on the Knock Control Adv. I make sure it's safe to drive it back to my tuner without towing that's it.
That's fine. Mine does it too, and it's supposed to. It isn't "knock", it's the total timing allowed to be added, is how I understand it.
Bottom line, it's normal, and supposed to do that.
I think the table name is misleading.

Bringin it back.

So I'm trying to work around this same situation right now, and I've been logging a trend that may just be some ridiculous thought in my head, but no ones mentioned it so far. It seems that when the fuel pump goes into 100% duty cycle there is a slight delay, as if to signify the increasing flow/pressure, and then the AFR's drop like a rock. I heard the phrase "pump shot" up above but once my AFR drops it stays there, even with constant throttle and steady peak boost.

So my question here is, what kind of pump is everyone running? I've got a walbro high pressure and I'm wondering if that isn't the problem.

(setup: FP18g, FMIC, TBE, cheap intake, walbro high pressure 255 fuel pump, stock injectors)

Bringin it back.

So I'm trying to work around this same situation right now, and I've been logging a trend that may just be some ridiculous thought in my head, but no ones mentioned it so far. It seems that when the fuel pump goes into 100% duty cycle there is a slight delay, as if to signify the increasing flow/pressure, and then the AFR's drop like a rock. I heard the phrase "pump shot" up above but once my AFR drops it stays there, even with constant throttle and steady peak boost.

So my question here is, what kind of pump is everyone running? I've got a walbro high pressure and I'm wondering if that isn't the problem.

(setup: FP18g, FMIC, TBE, cheap intake, walbro high pressure 255 fuel pump, stock injectors)

The fuel pump can't be the problem your Fuel pressure regulator controls your fuel pressure. Everyone runs the walbro 255 pump.
It is most likely a MAF scaling problem that needs to be adjusted for your intake. Or a boost leak. Boost leaks are notorious with FMIC.

Mine was, as hart said, MAF scaling..
Im rock-solid 11-11.2s under WOT now, since I've got my maf scaling 99% complete.

Well, let me say I feel pretty stupid. A couple weeks ago I accidentally created and fixed a boost leak, but just yesterday I found another blatant one. Don't know how I missed it, but it must have been the problem all along. 12's now under boost and the car feels a lot smoother.