Okay, after playing and spending a lot of hours on Evoscan, ECUflash, and Gunzo's base maps... and about a dozen trial/error versions of modding the 91oct map, I could say I'm getting really close to a ROM that I'm happy with.

Setup:
Evo X GSR 100% stock (switched back to stock panel filter)
Fuel: 91oct Costco Gas
ECU: Gunzo's 91oct base ROM w/ lesser timing adapted from his 90oct base map (I blame crappy CA gas =X)

Here is my 3rd gear (or was it 4th) pull from a flat deserted highway. Log played back using a customized Evoscan virtual gauge setup:

http://www.screentoaster.com/watch/s...eration_logged

note: peak boost was at 21.34 psi

I'm working hardcore to make further modifications on my current ROM. This includes:
1.) adapt a milder version of Hiboost's MIVEC maps which will be a little less aggressive than Gunzo's.
2.) improve boost just a little bit (factory tune still has a higher peak)
3.) modifying the fuel map to be to be a tad leaner than the 91oct base map, yet a little richer compared to the 93oct map (since I don't have a WB yet, I like it safe)

Thoughts on the 3 ideas above?

 

Can you please put the actual csv file in a zip folder and post it. While your gauges are cute, the actual log is much better to decipher.

And why are you tuning w/o a wideband? You said you want to play it safe, yet you are tuning w/o a wideband!!! As I always say: Tuning w/o a wideband is like fornicating w/o a condom ;-)

 

Attached are 2 back-to-back logs... I need to get those 3ks fixed

Car pulls strong (so much better than the factory tune).

=edit= The vid is just something that is easier to visualize especially for noobs like me :-)

 

^ LOL i know what you mean... but here are my thoughts why I went ahead and "tuned" my car even without a wideband:

1.) A lot of people have already tried all of Gunzo's maps. Those who had 100% stock Xs haven't reported any major issues. My car is all stock!
2.) I've studied and compared all the maps on the 3 ROMs and tried to understand the differences
3.) Any tweaks I made are within the "safe boundaries", as for timing I copied from the 90oct map since it basically had a max of 1 KS. The 91oct timing was a little high for our CA gas, I was getting 4ks sometimes.
4.) The fuel map on the 91oct doesn't deviate that much from the factory map, max is 1 point (theoritically comparable to running a stock X on factory tune w/ say an AEM intake). Since I went back to the stock filter, it should still yield relatively rich AFRs on the high load/throttle.

 

Thanks for sharing, but are those WOT logs? It doesn't look like it based on TPS value. Also, you should log to at least 7000 rpms, if not 7500. With 3 knock sum at 170 load / 4300 rpm at part throttle, I'd back off timing a bit more or richen it up. I've never seen my car knock at part throttle like that on 91. Like nj said, you really should get a wideband if you want to play it safe. What is your fuel injector scaling set at?

 

^ That log was taken last night with 91oct timing (slightly less in some areas)... the new version I'm going to log next will use the 90oct timing (factory timing is higher). I'm beginning to think it's the aggressive MIVEC map that gunzo uses, so next I'll try Hiboost's (with a few milder changes in some areas)

I wasn't ready for a WOT log yet since I saw those 3ks... will only do that once I get the ROM to zero in knocksums. I couldn't go over 6500 RPM since the speed limit on the highway was only 55mph (LOL), and it was nighttime, who knows a cop might be hiding somewhere.

As for the injectors, I haven't touched any other areas of Gunzo's maps except timing, fuel, & Mivec. His 91oct map has a default setting of 552, stock and generic90 is at 532. Should I set it to 532 also? Is lower safer in this case?

- edit - just checked his 93oct map, injector scaling is at 532... I wonder why he set the 91oct map with a higher value.

 

These logs are pointless. Your not going WOT. You need to log WOT from 2500 rpm to 7500 rpm.

 

If you want to be safer, set it to 532, which is the stock injector scaling value. When set higher, it tricks the car into thinking the injectors are capable of giving more fuel than it really can, so it runs leaner.

Also, as far as I know, Gunzo's MIVEC maps are just the JDM maps, and they should be MORE conservative than HiBoost's maps. If you're trying to be conservative, I'd stick the JDM mivec.

I'd still strongly suggest a wideband as your car is knocking even at part throttle.

 

^Got it to 532. My initial impression on Gunzo's modded JDM MIVEC map is that there are higher peaks in the lower RPMs and the changes don't look smooth. Hiboost's is perhaps more aggressive but the changes are smoother. I'm going to create my own version and try that out... based on Hiboost's but a bit patterned like stock... and a lot less aggressive.

UPDATE: I've also downgraded the base ROM to 90oct due to the occasional 1-2 KS under hard acceleration (~60% throttle). Costco gas is not helping. Will try 76 or Chevron next time.

 

Yeah... I'd avoid the Costco gas. Try 76, Chevron or Shell.

One of my buddies picked up a X and started tuning it recently, completely stock. I'm helping him out a little bit and I THINK I have a pretty good way of predicting A/F without a wideband. Only works for a completely stock car at stock boost levels though as I used wideband data posted taken from a completey stock car at stock boost levels.

Is there a way to log air mass flow rate in grams/second using Evoscan? That'd help a ton in calculating A/F. That should make it a relatively simple calculation actually.

 

Anybody got a link to Gunzo's 91 oct base map?

 

I read somewhere that the ECU target AFR map versus actual AFRs are predictably very close as long as the car is running on a completely stock setup (asumming you are not in the mountains and weather conditions are "normal") Putting on an aftermarket CBE/TBE or intake/filter will change the equation a lot, so that's probably when you will absolutely need a wideband to safely tune it.

funks: just google it using keywords "gunzo", ecuflash, & evoxforums

 

That is not correct on an Evo. On a Subaru, the target and actual AFR are close, but on an Evo that is not so.

Stop trying to justify tuning w/o a wideband. What you are doing is inexcusable. Buy a wideband before you start tuning.

 

I've noticed that it's only close up to about 5500rpms. After that, the car is leaner than the values in the fuel map.

The load value is somewhat close to wheel torque.... all ballpark.

 

Well said

Even if MY stock car had actual AFR close to map AFR, that does not mean someone else's stock car would be the same. Every car can be different, which is why a wideband is essential.