RomRaider

AZ1G202I MerpMod SD Flex Fuel GTX3076R GEN II WIP
Posting up my current rom from my 2008 STi daily driver. I've gotten pretty far from referencing others tunes on this website so I figure I should put mine out there as well. This is a full SD tune with tested flex capabilities up to about 40%. Big thanks to Jon7009 from continuely sharing great maps with the world. Even when I was tuning my bugeye 5 years ago his Carberry SD Base maps were a huge starting point for me. Important things to note if you are looking to use this tune as a basemap for yourself, would not recommend running the tune as is do to the major modifications on the car and potential differences to yours. All tuning has taken place at 5000ft plus elevation and the mod list is as follows.

Mods:
Perrin Rotated Kit w/ GTX 3076R Gen II 0.83 A/R T3
Perrin FMIC Kit
Invidia Q300 Catback
CNT Racing V3 Even Length Headers
Air Pump Delete
TGV Deletes
MAC 3 Port BCS
AEM 3.5 Bar MAP Sensor
GM IAT Post Intercooler
Stock Fuel Rails Modified for Parallel Fueling
Stock FPR
Cobb ID1300-XDS
Stock EJ257 DAVCS Longblock

General Tune Notes:
AVCS values have been modified for more torque and spoolup. Requested Torque tables modified to try to aid percieved spoolup. Turbo takes a lot more to get going at this altitude. Timing is kept conservative due to running primarily on suspect 91 pump gas. Speed density smoothing zerod for more responsiveness and idle stability values slightly altered to make catching idle smoother. VE Table needs a lot of work I suspect I have a mechanical issue with fuel pressure causing some irregularities, but close loop works well enough I don't care too much. Pops and bangs on intelligent drive mode, sounds awful through the well muffled exhaust. E85 is largely unteseted. Lower blends of up to about 40% worked great, albeit the power gains are marginal, I will test higher content when the weather warms up. Lots of the cars time is spent above 8k feet and under 0 degrees F so I would rather have the lubricity and lower fuel dilution of gasoline in the winter.

Experimental Defs:
Many definitions have been added to the AZ1G202I.xml file for ECUFlash. 32BITBASE has had a few modifications for scalings for the tables added to AZ1G202I. Many addresses were found from snooping through the forums as well as a few from ScoobyRom. Rev hang has been addressed with the Decel Adder B table. Its A counterpart doesn't seem to affect anything atleast on my car. Idle stability was greatly improved with the addition of Low Pulse Width Compensation Values for the ID1300s. Also the MPG display on the dash has been modified to be accurate to calculated MPG with 0.3MPG error either way. This is necassary after changing injector scalings and will be specific in this case to the ID1300s. These are all included with the rom itself in the attatched zip.

Feedback/Questions are welcomed. The rom is attatched along with the modified definition files with the extended tables.

There is a lot to cover here. I guess I'll cover them in sections. First we'll talk about barometric comps and IAT comps.

If you never travel outside of a small normal window of barometric pressure, you'd have no reason to adjust the SD Atmospheric pressure comp, but ideally you would never need to worry about larger jumps with it set up right. your VE table would be good across all barometric pressure ranges.

airmass is calculated by

DISPLACEMENT * MANIFOLD PRESSURE ABSOLUTE * VE% CELL * SD MULTIPLIER, then other comps like for IAT.

The SD multiplier is grams of airmass per psia per revolution

Nowhere in that is barometric pressure.

The VE table is listed in the common definitions as "Manifold Relative Pressure (Sea Level)" but what this really means is MAP - 14.7. The VE table is actually defined as absolute pressure, torr or mmHg specifically, but opensource people decided to define it as MRP (sea level)

So when you're not at 14.7 psi and you're logging Boost (MAP - BARO) and then you adjust your VE table, you're not actually touching the place in the VE table where your logged boost was actually at.

10 psi at sea level, as the VE table would show, is 24.7 PSIa.
10 psi at 13 psi of baro is 23 PSIa. you'd be adjusting by an error of 1.7 PSi.

better to convert the table to read in MAP, datalog MAP, and forget datalogging boost. just datalog boost error.

guess what? yep. target boost isn't actually a 'boost' or gauge relative table. weeeeee! it's an absolute pressure value in the z-axis, just subtracted by 14.7 to give you the perception it's a gauge relative table. nope.

so, what I do is think of the boost target as an absolute pressure target at the outlet side of the compressor, with the baro being what would be something like the absolute pressure at the inlet of the compressor, and from that we have a pressure ratio.

we set up a certain pressure ratio, and then adjust the 'target boost' multiplier and offset so that we effectively get the same compressor pressure ratio targets at all barometric pressure.

if the math is good, then you never exactly see at your boost gauge what is in the boost table, but your boost error could always be the same, very minimal, across all baro, and your turbo would be ideally, nearly all the time at the same pressure ratio at the same throttle/rpm load point across all baro.

then you only need MAP and Boost Error, don't need to worry if your boost gauge matches your 'target boost table'.

as a secondary note, "robbing peter to pay paul" by using the VE table to comp for whatever lack of compensation should be present against load elsewhere (baro / IAT comp) then you cause load drift whenever you go outside of whatever temperature average and baro average you tuned your VE table to. if you're tuning your VE table in different ranges with them being tuned with different IATs and different baro, then each of the areas of the VE table would have been 'centered' to those IAT/baro ranges. best to comp everything properly and then you just use MAP and RPM for the VE table, not worrying about temperature swings or traveling hitherto.

TL;DR - Careful about using the logged boost parameter to adjust any table has an axis type or table value of 'boost' or gauge relative, because the tables are actually in absolute pressure values and it's just defined for convenience of the novice.

IAG Thursday's latest video on youtube is a fantastic discussion on AVCS. I'll go over some of that here.

0 on the Intake AVCS table is centerline for the intake cam, as non-avcs cams would be.

17 degrees retarded in the Exhaust AVCS cam table is centerline for the exhaust cam, as non-avcs exhaust cams would be.

0 degrees in the exhaust cam table means you're advanced, as such, is when you disable avcs control.

exhaust retard can help spool and torque but only in the lower rpms. on the reverse, if we advance the exhaust cam at early rpms like leaving it unplugged, we lose boost and torque in the early rpms.

mostly the exhaust cam is used for emissions from the oem standpoint, and not so much for power.

I prefer to basically flatten it out, leaving just the idle/off-idle area where it is 0 (because there isn't a ton of oil pressure to be moving the cam) and because it's fine for emissions/mileage, and make the rest of the table 17.

this makes all your cruise island for efficiency, low rpm ramp in, to high rpm ramp out for cam timing happen in intake avcs, so effectively you could run the same intake avcs table for both DAVCS and SAVCS.

That's where you start, ideally, IMO, and then you add or subtract from exhaust avcs for reasons of either increasing top end power (advance, reduce the retard value) or add spool (retard, increase value)

if we look at both your avcs tables at full load at 6000 rpm, you have 22 degrees exhaust retard, and 0 degrees intake advance. this gives you 5 degrees extra overlap over centerline(s), but entirely from the exhaust cam perspective. if overlap at higher rpm is good, it wouldn't be from the exhaust cam, because that limits higher rpms and is better for torque/spool at low rpms, right? it would be better to have 5 degrees of intake avcs and 17 degrees of exhaust avcs, this way you still get the same overlap, but advanced by 5 degrees for both cams.

if you retard the exhaust cam, you get overlap. advancing the intake cam, also overlap. straight to overlap. (insert portlandia meme here)

so really, IMO, better to start with semi aggro intake avcs and a softer exhaust avcs.

It's the man himself! Thanks so much for the reply. Since this post I've since taken the car to the beach and back, from Colorado, and I've switched the tune over to DimeMod. It was very rough at first as all my experience thus far has been Carberry/Merpmod/Haltech. Haltech is cheating its so easy compared to open source. Back on topic I've implemented your suggested AVCS Changes into my Dimemod tune and the difference is night and day. My VE map is looking more normal too; I found my fuel pump was dying when I did the Merpmod table. Car is overall very happy, but I'm busy working out more bugs now. I'm not sure where posting Dime Roms lands on the rules section. It's all self tuned but it seems Dimemod isn't super common to find on the internet . Also having major issues cold starting on E80 which is to be expected with the more limited flex tables compared to my last opensource tune up here on Carberry. That car had no issues down to even -10F. I've received word from Dime that he is working on adding more cold start focused E85 tables which would be invaluable in the mountains. If posting my new Dime Rom with definitions is allowed I'll throw it up here shortly once I confirm that.