RomRaider

updates for leaning out closed loop for ECT comp, base AFR + load comp for CL, and other compensations

Hey Jon, off-topic question for ya. Since you've been using MLV much longer than I, is it possible to use VE Analyser with our tunes or should I do it the old fashioned way with some histograms?

I'm using an excel spreadsheet that takes the VE table as it was in the rom, and you paste the closed loop correction data from histogram with calculated field for total trim correction and filters

copy / paste CL trim adjustment
copy / paste OL trim adjustment

output table appears on bottom

Once I polish it up a little I'll post it along with some suggestions for calculated fields and filters.

I modified these values with those of a WRX BASE MAP. But once it has a normal temperature, the AFR sometimes returns me 15.5. And in cold starts, the revolutions range between 1000 and 2000 rpm.

I have a Forester S-Turbo MY 2002 EJ205 (A4SE700J) with TD04, oldschool TGV delete, 2.5 "catless downpipe, 440cc injectors, forged pistons / rods, 4,444 final drive, walbro 255, NGK iridium.

I have been with this topic for a year. I am aware of what I do, with little knowledge. Here in Argentina, there are few who know about subaru tunning. My fear is leaving him with someone who ends up exploiting everything!

I think just i need a Stage 1 map...(wrx 2002map)...

How would one enable TGV control on this rom? thanks for sharing.

Hey lads just wondering what i would need to change if i still have the stock 02 wrx injectors, just got into tunein me car. If anyone knows the values for all the tables to be changed that would be amazing. Cheers

well, up near the top of the thread I mention the 9 or so tables you would change for injectors, but I also made this base and was able to test it with a couple vehicles and fine tune the target and wastegate duty tables.

91+ AKI
02/03 USDM WRX EJ205 / USDM S20 (small port) HEADS
02 - 05 WRX 5 speed (3.90 f.diff, 1.1 transfer ratio, 3.54 r.diff)

stock location (manifold) upstream planar lambda sensor (that's the specific type of o2 sensor for you non-nerds)
stock with TGVs
stock TD04L-13T turbo
stock BCS / pill
stock MAP sensor
stock bypass valve
stock MAF / intake
stock injectors
stock fuel pump / or with walbro 255LPH (GSS-342)
stock black coil packs
stock heat range plugs (NGK #6, 0.028")
stock evap system
stock IAT sensor (the visible thermister when you look at the MAF sensor body)

+turbo inlet (like Perrin, Mamba, etc.)
+catless up-pipe
+catless down-pipe
+STi TMIC / IC shroud +/- STi scoop

-rear o2 stuff disabled

*Knock-CEL enabled when either FLKC or FKBC happens at greater than -2.81 degrees
*Launch Control with defogger switch

hey Jon, I hadn't tried this out just as of yet, but for anyone learning VE tuning, is there a way to easily calculate additional efficiency when adding more boost? Would it just go about adding x% values to your VE map?

datalog fuel trims, and your basic inputs:

short term fuel trim (A/F Correction #1)
long term fuel trim (A/F Learning #1)
closed loop / open loop status
injector duty cycle
injector #1 pulsewidth
fueling final base
wideband
throttle percent
manifold absolute pressure
engine speed

duty cycle is used as a determination for decel fuel cut, where fueling trims are ignored
pulsewidth is used to see if you're near or hitting the minimum ipw

also if you can represent your datalog in graphs/histograms, then y axis of IDC and x axis of MAPxRPM (times) can show you how well your injector calibration is holding up or if your closed loop fuel trims are correcting for scalar issues, etc.

short term fuel trim comes as a percentage, as does long term fuel trim, so for closed loop, you make each one into a factor and multiply them against each-other and the VE cell that you are at for the time.

requested fueling final base and wideband are used when in open loop to determine a factor for correction. it's Wideband AFR / FFB AFR, or WB Lambda / FFB lambda, etc. if you're requesting say 11.2:1 and you get 12:1, then your correction is 12/11.2 and you get a factor of 1.0714~ to that VE cell.

using something like Megalogviewer HD with it's calculated fields, histograms, etc. is where it's at. sure, you can get a spreadsheet or two and do the work that way, but it's more time consuming and you certainly can't just as easily sort through a hundred+ MB of logs

more than that, you want to have some threshold for too much throttle change either direction, to rule out transitional fueling states that the VE table would be less responsible for at that point. filters are great.

if you're working with smaller logs, something like Smentuner can be used with excel to get you familiar with what's going on, but if you can afford $40, then MLVHD is like, the ****

Damn dude, thanks for the master class in this...


It snowed last nigbt in toronto, so its gonna be a few days until logs, but i lifted an sd table, used your blending polynomial table, and enabled blending 80@14.8 psi atmosphere, then 100% at 20psi until 33psi and watched afr targets, then scaled out the sd table until i got in line with my open loop fueling values. They are bang on across the board.

Ill start logging the parameters you outlined.

Hi Jon,
could you please elaborate more on this?As regards to tuning the fuel resume minimum injector pulse width,decelaration fuel cut,if you have low pulsewidth enabled and running 1000cc injectors.I take it the minimum pulsewidth 1.952ms is for stock blue wrx injectors and maf?so if running full sd,different injectors,how do you go adjusting this table?old injector/new injector correction factor?
Quote: duty cycle is used as a determination for decel fuel cut, where fueling trims are ignored
pulsewidth is used to see if you're near or hitting the minimum ipw..

decel fuel cut resume pulsewidth is too rich for our mods, first it needs to be lowered, and then it's a factor of old injector / new injector. DFC pulsewidth resume is not a minimum pulsewidth.

low pulsewidth comp is a correction to the lower end of the flow of the injector, to bring the short pulsewidth area up (or down!) to match the linear flow curve of the rest of the injector

minimum pulsewidth is the hard minimum that the injector can do in injection time. Say 0.380ms right, if your vehicle thinks it needs less fuel, it can't, as it can only ride that minimum limit.

When duty cycle is 0%, decel fuel cut happens. you don't want to be averaging zeros from fueling trims with actual data, or using lean wideband when there is no fueling or spark, for VE corrections.

Hi Thanks for the elaboration,though am still not understanding as below,
1.( decel fuel cut resume pulsewidth is too rich for our mods, first it needs to be lowered),how much lower?what informs how low you can go?
2. If its not a lot to ask,could you please give an example of fuel cut resume pulsewidth working with any size injector to arrive at the fuel cut resume pulsewidth to use for that size injector,so that i can get the math behind it?
Regards

in the rom I posted.

divide by old displacement, multiply by new displacement
multiply by old injector flow, divide by new injector flow

so for example:

1018.46cc/min for ethanol free pump gas on ID1050x on 1994cc engine:
resume pulsewidth: 0.752ms

812.5cc/min on E10 pump on modded densos on 1994cc engine:
resume pulsewidth: 0.716ms

that's on the low side

ride the engine at 0% throttle in like second gear, and wait for the rpm to decel enough (and the slower, the better) and when DFC resume pulsewidth happens, you'll know. you'll know ESPECIALLY if you have larger injectors, because the old value is different. if it's rich enough you'll get a nice POP.

Hi Jon, first thank you so much for this base map, I was totally lost with Carberry because all the conversion stuff, with this base map seems to be an excellent start to learn and tune with safe using SD, my car it's a wrx 2003 ASDM I just get this car 1 month ago and I made a little upgrades:

1. Up and Down-pipe catless.
2. STi TMIC
3. TGV Delete

I have the stock injectors, stock intake, stock turbo inlet and stock turbo, I notice that this rom have a aftermarket turbo inlet can this affect me? And other question how can I rise the turbo target to 18 psi? Thanks a lot again and in advance for any help.