Tuning General Discussion [split]

[Hotwire]

Good tuning work there for stock Mark and no doubt much more enjoyable to drive. LOL to "Port Hacking" as it sounds like a perfect place to hack into your ecu.

There's been a lot of hacking going on there lately. I thought this was going to be a lot easier than this to sort things out but then you find what tables effect other tables and you end up going backwards at times. Other ECUs are incredibly simple by comparison none of this level of detail in relation to learning, knock control and failsafes. Even the previous non CAN ECU's seem a fair bit simpler plus the 2L motor has proper pistons albeit at the expense of low down torque. All Fun
When I get some more time I'll run the car with the later maps and then do another back to stock using the same settings day etc and get some prettier graphs

Interesting you post this Mark. I've been reading and learning about ECU tuning, and logging via RomRaider for developing a tune for my 04 GT Wagon (with pre-can ecu & EJ20) and am wondering if you would mind helping out?!

If I send you logs of runs, can you help me with which modficiations to make to the tables etc? sort of remote tuning? I can get access to a local dyno also if need be (although its at ~$150 per hour would rather road tune if I can). Im very interested to learn how to tune, and you certainly seem to know your stuff.

Cheers
Lee

[parso_rex]

Don't get your hopes up too much. There are lots of guys round with way more experience in these than me. Like a few others I have an electronics IT background and it all seems quite logical and easy to understand. Personally I think the way to do it would be to see if there is a similar modded file around somewhere by searching romraider or Nasioc's forums and see what they have done and why. Probably the best option would be to have someone tune the car then you can see what they have done and take it from there. You'd need the obligatory LC1 and second bung in your dump to suit of course, you probably already have it. Your in SA which kind of limits you though

If we do a car with whatever ECU it comes in and goes straight on the dyno and you see what its up too. This way you have a known base to work from. If a car been "tuned" elsewhere you might be in for some real fun. You can't see this on the street easily nor can you load things up in the same way so thats why I like dynos. Having said that some oldhands are masters of roadtunes and actually get quite close. They end up finishing with a dyno after the road tune an then maybe back again - taking days !!. You'd always hav e to finish with a roadtune to load things in a similar way to normal
The 04 if I'm not mistaken has AVCS on the exhaust as well hmm never touched this in my life so far. I will see if I can find that doco that someone compiled on Nasioc "guide to tuning" or similar... here it is http://web.njit.edu/~trb2/WRX/download/

Mark

[Hotwire]

Thanks for the link Mark. Will be sure to have a good read

Where I have access to the dyno, the guy who owns it is rewnowned for his EVO/Mitsi tuning, and he said hes is happy to look at tuning a Subaru. Based on your knowledge, would he have any trouble tuning Subaru? (with the only exception I can see being the AVCS on both intake and exhaust as only EVO IX have variable valve timing on intake, not exhaust)

Cheers
Lee

[parso_rex]

I think the EVO's are easier myself there isn't as much to modify. You can see this in ECUFLASH which has the ability to edit the Mitsubishi ROMS as well. I have a whole heap of Stock roms of both brands plus a few modded ones - no modded ones of yours though
Only the EVO9s + have AVCS and In general the Mitsu donks are a bit different you don't need as much timing plus they run bigger cam profiles Their cast iron Block (pre EVO10) are very ridgid and have proper rods and pistons.
If the guy can tune EVOs and other cars properly its just a matter of learning how to use the software and flashing progs - all of which takes a fair bit of time but if he wants to have the ability to tune them in the future then its a good opportunity for him to learn.

[Hotwire]

I think the EVO's are easier myself there isn't as much to modify. You can see this in ECUFLASH which has the ability to edit the Mitsubishi ROMS as well. I have a whole heap of Stock roms of both brands plus a few modded ones - no modded ones of yours though
Only the EVO9s + have AVCS and In general the Mitsu donks are a bit different you don't need as much timing plus they run bigger cam profiles Their cast iron Block (pre EVO10) are very ridgid and have proper rods and pistons.
If the guy can tune EVOs and other cars properly its just a matter of learning how to use the software and flashing progs - all of which takes a fair bit of time but if he wants to have the ability to tune them in the future then its a good opportunity for him to learn.

I have begun reading through the tuning document and its great reading so far. One intetersting point I've picked up from the first few pages is the turbo efficiency maps, and keeping within the efficiency range of the turbo & intercooler. I've since searched the net for ages trying to find a compressor efficiency map for the VF38 with no luck, let alone what the efficiency of the stock intercooler would be. If you have any idea where I might find this info would be great.

Thanks
Lee

[parso_rex]

I think the EVO's are easier myself there isn't as much to modify. You can see this in ECUFLASH which has the ability to edit the Mitsubishi ROMS as well. I have a whole heap of Stock roms of both brands plus a few modded ones - no modded ones of yours though
Only the EVO9s + have AVCS and In general the Mitsu donks are a bit different you don't need as much timing plus they run bigger cam profiles Their cast iron Block (pre EVO10) are very ridgid and have proper rods and pistons.
If the guy can tune EVOs and other cars properly its just a matter of learning how to use the software and flashing progs - all of which takes a fair bit of time but if he wants to have the ability to tune them in the future then its a good opportunity for him to learn.

I have begun reading through the tuning document and its great reading so far. One intetersting point I've picked up from the first few pages is the turbo efficiency maps, and keeping within the efficiency range of the turbo & intercooler. I've since searched the net for ages trying to find a compressor efficiency map for the VF38 with no luck, let alone what the efficiency of the stock intercooler would be. If you have any idea where I might find this info would be great.

Thanks
Lee

If you into the turbo stuff and have a bit of free time check out the Garrett Technical documents doc 2&3 is where its all happening and quite technical in parts.
http://www.turbobygarrett.com/turbobyga ... enter.html

I have some data about the VF series and others from the Rexnet site only basic though but it may be all you need

VF22
This turbo has the highest output potential of all of the IHI VF series turbos and is the best choice for those who are looking for loads of top end power. The top end power however, does not come without a cost. The VF22 spools significantly slower than the rest of the IHI models due to the larger P20 exhaust housing and is much less suited for daily driving than some of the other models. Although the largest VF series turbo, the VF22 is not quite optimal for stroked engines or those who wish to run more than 20PSI of boost.

VF23
This turbo is considered a great all-around turbo. Like the VF22 it utilizes the largest P20 exhaust housing. This housing is mated with a smaller compressor housing of the of the VF24. This turbo is considered optimal in applications with range from mild to slightly wild. It does not have the same top end power of the VF22, but spools up significantly quicker.

VF24
This turbo shares its compressor housing with the VF23 however, this housing is mated with a smaller (P18) exhaust side. The smaller characteristics of this turbo allow it to provide ample bottom end power and quick spool. This turbo is very popular for Imprezas with automatic transmissions and Group N rally cars.

VF28
This turbo came standard on the STi Version 5. In terms of overall size, it is smaller than the VF22, VF30 and VF34, and about same size as the VF23.

VF29
This Turbo is nearly identical to the VF24, with the same compressor and exhaust housings. However the compressor wheel in the VF29 is has been changed slightly. The changes made to the compressor wheel in this model are generally viewed as improvements, and as such this unit is typically chosen over the VF24.

VF30
The VF30 is commonly considered the best bang for the buck turbo in the IHI VF series line. A relatively new model the VF30 features the same exhaust housing as the VF24 but a larger compressor side similar to the VF22. The combination of these two parts results in increased output potential without the lag associated with the VF22. Although it doesn't offer the top end supremacy of the VF22, the VF30 is a great compromise between these unit and the quicker spooling models.

VF34
The VF34 is nearly identical to the VF30, with the same exhaust housing and compressor. However the VF34 goes back to the ball bearing design, and in doing so achieves full boost approximately 500RPM sooner than the comparable VF30. The VF34 is the most recent IHI design and as such costs slightly more than its counterpart. Top end performance and maximum output are identical to the 30.

VF35
The VF35 has identical internals as the VF30 and it uses divided thrust bearings. However, the exhaust housing is a P15 which means this turbo will have fantastic spool characteristics. This turbo is standard on the new WRX Type RA. LIMITED SUPPLY.

VF36
Roller bearing version of the twin scroll VF37, also has a titanium turbine and shaft for even quicker spool. Same compressor housing as VF30/34, however twin scroll P25 exhaust housing provides slightly better top end output due to reduced exhaust pulse interference. This turbo is good for 400HP and used on JDM STI Spec C from 2003 onwards.

VF37 (thrust bearing)
Enter the age of twin scroll IHI turbos. Same compressor housing as VF30/34, however has a new twin scroll P25 exhaust housing that provides slightly better top end output due to reduced exhaust pulse interference. Twin scroll also provides better spool up for improved low down response over the VF30/34. This turbo is good for 400HP and used on JDM STI from 2003 onwards.

VF38
Twin scroll turbo with titanium turbine and shaft. Smaller compressor housing than VF36/VF37 provides tremendous spool up capabilities but less top end than VF36/37. The spool capabilities of this turbo are demonstrated on the JDM Legacy GT, which reaches peak torque at 2400RPM.

VF39
Single scroll turbo used on USDM STI and latest 2.5L STIs released internationally. Smaller than VF30/VF34.

VF42
Exclusive turbo to the S203/S204 models, this features a twin scroll design with a slightly larger compressor than the VF36/37 turbos and different turbine design (more blades). The VF42 is a roller-bearing turbo and is likely of similar size to the VF22 turbo, but with twin scroll exhaust housing for faster spool and superior top end performance due to reduced exhaust pulse interference.

PE 1818 The 1818 supposedly comes on boost extremely fast...which make it great for road racing or aggressive driving (remember to be responsible :^). The 1818 maxes out somewhere between 350-375 hp.

PE 1820 The 1820 is larger than the 1818 and is capable of running more boost, but at the expense of more turbo lag. Expect it not attain full boost until you have passed 4000 rpm. However, it can be tuned to 400+ hp.



IHI & Garret Turbo Specs
Model Type Compressor Inlet Diameter Compressor Housing Nominal Turbine Housing Turbine Approx A/R Compressor Wheel Blades Turbine Wheel Blades Waste Gate Actuator Turbine Casting ID & Batch
VF22 Roller Bearing 48.5 A 9.4 PZ20-H 94001 0.71 5 11 C395 H, H4
VF 23 Roller Bearing 46.7 B 9.4 PZ20-H 94001 0.71 6 11 C395 8psi TBA
VF 24 Roller Bearing 46.7 B 9.4 PZ18-H 94001 0.63 6 11 C418 8psi H, S4, 66
VF 30 Bronze Bearing 47.9 C 9.4 PZ18-H 94001 0.63 6 11 C484 H, S4, IC F55
VF 34 Roller Bearing

Garrett 400 Ball Bearing 53 - - Outlet Dia 51.5 - 7 9 11 psi
Garrett 450-500 Ball Bearing 53 Outlet Dia 46.9 - 7 10 14 psi




Turbo Type ----------- Approx flow @ pressure
Stock Turbo ---------- 360 CFM at 14.7 PSI
IHI VF 25 ------------- 370 CFM at 14.7 PSI
IHI VF 26 ------------- 390 CFM at 14.7 PSI
T3 60 trim ----------- 400 CFM at 14.7 PSI
IHI VF 27 ------------- 400 CFM at 14.7 PSI
IHI VF 24/28/29 ----- 410 CFM at 14.7 PSI
========= 422 CFM max flow for a 2 Liter at .85 VE pressure ratio 2.0 (14.7 PSI) 7000 RPM =======
IHI VF 23 ------------- 423 CFM at 14.7 PSI
FP STOCK HYBRID -- 430 CFM at 14.7 PSI
IHI VF-30 ------------- 435 CFM at 14.7 PSI
SR 30 ----------------- 435 CFM at 14.7 PSI
IHI VF-22 ------------ 440 CFM at 14.7 PSI
T04E 40 trim -------- 460 CFM at 14.7 PSI
========= 464 CFM max flow for a 2.2 Liter at .85 VE pressure ratio 2.0 (14.7 PSI) 7000 rpm =======
PE1818 -------------- 490 CFM at 14.7 PSI
Small 16G ------------ 505 CFM at 14.7 PSI
ION Spec (stg 0) --- 525 CFM at 14.7 PSI
========= 526 CFM max flow for a 2.5 Liter at .85 VE pressure ratio 2.0 (14.7 PSI) 7000 RPM =======
Large 16G ----------- 550 CFM at 14.7 PSI
SR 40 ----------------- 595 CFM at 14.7 PSI
18G ------------------- 600 CFM at 14.7 PSI
PE 1820 -------------- 630 CFM at 14.7 PSI
20G ------------------ 650 CFM at 14.7 PSI
SR 50 ---------------- 710 CFM at 14.7 PSI
GT-30 ---------------- 725 CFM at 14.7 PSI
60-1 ----------------- 725 CFM at 14.7 PSI
GT-35R -------------- 820 CFM at 14.7 PSI
T72 ------------------ 920 CFM at 14.7 PSI <--- Note you would have to spin a 2.0 L engine at about 14,000 rpm to flow this much air.
IHI VF 25 ----------- 395 CFM at 18 PSI
IHI VF 26 ----------- 400 CFM at 18 PSI
T3 60 trim ---------- 410 CFM at 20 PSI
IHI VF 27 ----------- 420 CFM at 18 PSI
IHI VF 24/28/29 -- 425 CFM at 18 PSI
IHI VF 23 ----------- 430 CFM at 18 PSI
IHI VF-30 ----------- 460 CFM at 18.0 PSI
AVO 320HP -------- 465 CFM at 17.5 PSI
T04E 40 trim ------ 465 CFM at 22 PSI
FP STOCK HYBRID- 490 CFM at 18.0 PSI
IHI VF-22 ---------- 490 CFM at 18.0 PSI
SR 30 --------------- 490 CFM at 22 PSI
Small 16G ---------- 490 CFM at 22 PSI
ION Spec (stg 0) - 500 CFM at 19 PSI
PE1818 ------------ 515 CFM at 22 PSI
Large 16G --------- 520 CFM at 22 PSI
========= 526 CFM max flow for a 2 Liter at .85 VE pressure ratio 2.5 (22 PSI) 7000 rpm =======
========= 578 CFM max flow for a 2.2 Liter at .85 VE pressure ratio 2.5 (22 PSI) 7000 rpm =======
HKS GT2835 ------- 580 CFM at 22 PSI 400 hp
MRT 400 ------------ 580 CFM at 16 PSI
AVO 400HP -------- 580 CFM at 17.5 PSI
MRT 450 ------------ 650 CFM at 19 PSI
AVO 450HP -------- 650 CFM at 20.0 PSI
SR 40 ---------------- 650 CFM at 22 PSI
========= 658 CFM max flow for a 2.5 Liter at .85 VE pressure ratio 2.5 (22 PSI) 7000 rpm =======
HKS GT3037 ------ 670 CFM at 22 PSI 460 hp
PE 1820 ----------- 680 CFM at 22 PSI
20G ---------------- 695 CFM at 20.0 PSI
HKS GT3040 ----- 710 CFM at 22 PSI 490 hp
AVO 500HP ------ 770 CFM at 22 PSI
SR 50 ------------- 770 CFM at 22 PSI
GT-30 ------------- 790 CFM at 22 PSI
60-1 --------------- 800 CFM at 22 PSI
HKS GT3240 ----- 830 CFM at 22 PSI 570 hp
GT-35R ----------- 880 CFM at 22 PSI
T72 --------------- 1000 CFM at 22 PSI <--- note you would have to run a 2.0 L engine at >40 PSI boost to flow this much air
Conversions used where there was control over conversion factors:
1 HP approx equals 1.45 CFM
1 CFM approx equals 0.0745 lb of air/min
0.108 Lb/min approx equals 1 hp
1 Meter cubed/sec = 35.314 CFS = 2118.867 CFM
1 KG/sec = 132 lbs/min approx equals 1771.812 CFM
power coversions:
1 PS = 0.9859 HP = 75 Kgf m/sec
1.3405 HP = 1 KW
1 HP = 746 watts

That should keep you busy for a while LOL

[Hotwire]

OK so I've read the the NASIOC tuning guide and Im thinking the following - tell me if im on the right track:

Following the steps of the tuning guide:
1) Establish Air and Fuel relationship
- I dont think I need to do this as HKS Panel is only intake mod
2) Tune boost
- Will need to do this as I want to potentially increase boost in Low end by couple of PSI. Really would like to see what others have done for a GT here, but not looking good....
3) Tune Fuel
- Take it this is the part that really needs dyno work
4) Tune timing
- as above
5) Tune AVCS
- Believe this is where my best gains for low end would be - again don't want to start from scratch but see what others are doing.
6) Retune fuel and timing
- More dyno time

So in summary, would be far better if I could start with another MAP, load it and log it on the road using learning view/RomRaider logger. Work out where the map needs adjustment and go from there. This would minimise my dyno time in a big way, but trouble is getting a tuned map for the 2.0..............

[smythie]

1) Establish Air and Fuel relationship
- I dont think I need to do this as HKS Panel is only intake mod

Not a tuner myself but from every comment I have heard/read from various tuners I reckon you do need to find out just what air/fuel ratios your car is running and look at exhaust gas analysis to give you the best idea of what state your current level of tune is at. It will probably end up being part of #3 but it is still necessary base data I think

[Hotwire]

1) Establish Air and Fuel relationship
- I dont think I need to do this as HKS Panel is only intake mod

Not a tuner myself but from every comment I have heard/read from various tuners I reckon you do need to find out just what air/fuel ratios your car is running and look at exhaust gas analysis to give you the best idea of what state your current level of tune is at. It will probably end up being part of #3 but it is still necessary base data I think

In the tuning docco though, "Establish Air and Fuel relationship" discusses the scaling of injectors and scaling of the MAF for if you change the fuel delivery (new pump, injectors etc) or if you alter the inlet tract, ie. pod in guard etc. So your right that addressing the AFR's is done in #3 & #4, but I haven't changed the air or fuel delivery components

[smythie]

fair enough

[parso_rex]

You don't want to start messing with the MAF scaling if you can help it and a panel filter aint gonna affect that much at all - if any. I know some people claim it can but I can't believe that.
I'd leave AVCS alone at first, maybe on that model its not so weirdly setup anyway plus you have exhaust AVCS too. Just leave it and that can be round 2

[Hotwire]

Hi Mark

Can you shed some light on this for me - today I finally got RomRaider connected for my drive home from work. I was logging the drive including sitting in traffic and a couple of WOT runs (both up hill), and it felt like it was hesitating under load heading up the hill. So I got Learning View working and came up with the following:

LearningView_SS_4-08-2009 5 17 03 PM.jpg (81.62 KiB) Viewed 7759 times

Now this seems rather concerning to me that a completely stock car, with stock MAP and running BP Ultimate fuel is getting knock values causing IAM to be less than 1!! The hesitation I mentioned I was feeling before I read the ROM so thinking something is not right

Now i did read the ECU on Sunday, and the car has probably been driven a total of about 1 hour since then, but I would've thought it should have learnt knock values etc by now - or is it still learning since the reset and thus why the IAM is less than 1?

[Hotwire]

oh, and to add - looking at the logs, the majority of knock correction logged in RomRaider is +ve, only 30 counts of -1 in 3000 log data points, all at partial throttle never WOT - does this mean its still learning and not to worry about?

[parso_rex]

Thats not good, if you look at the fine learning knock correction table theres knock in quite a few areas and it would seem enough to drop your IAM too. Now what is the knock and is it even possible that its false knocking ?. Try the next tank on Shell 98 for some reason I like this fuel now, I used to swear by BP but the shell has been consistently good for tuning so I've switched teams for the time being.
False knocking: some guys get some knock when they get off and back on the gas thats acceptable especially with the 2.5L but yours is in all different ranges. My mates STI was driving him nuts with timing being retarded due to knock dropping his IAM down. He thought the knock sensor was faulty and ordered another form the states. I kept telling him to check his exhaust.months later It turns out that his front exhaust hanger wasn't attached and was bashing on the gearbox . I now have the same issue on my STI where the rotated mounts dump pipe has broken its mount too. The autronic doesn't give a rats about knock (with out the optional full tilt knock board) but I have a knocklink monitoring box and its been going wild at times due to the rattling mount. I should fix that I suppose
Maybe its not this at all but its worth looking at before you think about the tuning side of things especially as its stock
EDIT: turbo heatshields rattle too if they are cracked

[parso_rex]

Hang on a sec so did you read this learning view capture after you did a reset ?? if so it might well be learning. reset the thing again and just take it easy with a few pulls above 3700 or so but not flat to the floor and it will bring up the IAM quicker. But if its getting knock for whatever reason it may never see 1