AFR Tuning Tips?
 

Hey all, can anyone give me some tips on how to tune AFR? I know how to edit the map, but I'm wondering what ideal AFR's I should be tuning for.

Right now, I'm trying to get 12.5:1 during spool up (2500-3500 rpm's), 11.5:1 at peak boost (3500-5000 rpm's), tapering down to 11:1 at redline.

However, with how rich the stock map is under low load (100) at 2000 and 2500 rpm's, I'm running mid to high 13 AFR's until 3000 rpm's when I'm already building a bit of boost. I richened it up quite a bit, but I'm just not sure HOW rich I should make it. It seems that if I want to really run 12:5:1 during spool, I will have to significantly richen up the low load cells at 2500 rpm's (leaving a big jump from 2000 rpms which is still 14.7 to 2500 rpm's which is in the 12's on the map).

I'd REALLY appreciate any advice anyone can give me on this. I spent a lot of time trying to tune the AFR and it's been quite an experience. Thanks GUNZO for the base map to starts with! It's been a huge help! {thumbup} And thanks TEPHRA for making all of this possible with your logger :)

The leaner you can keep it while spooling up the turbo without hitting knock of course the better, although that still has to be within reason. Making the maps richer in the 100-180 load cells seems to counter that strategy of using a lean spool setup though. I'm assuming you are datalogging with a wideband, overall your ideas to aim for 11.5 at torque peak and 11.0 at redline seem perfectly fine, although you may be able to keep 11.5 all the way through the rev range. Every car reacts a little different but I've read that some cars have been tuned leaner with good success by dialing back the timing slightly and ensuring no knock.

If you plan on tracking the car at a road course then being slightly richer is a good idea since you need to make sure the car is tuned safe after 20-30 min of hammering on it. You will also find that the JDM Mivec maps may give you a better spoolup you are seaking requiring minimal changes to the main fuel and timing maps. Basically retarding the exhaust side and advancing the intake during the spoolup load cells can make a huge difference. More extreme changes there might require richer fueling or less timing. I'm still testing the changes I made to my maps but will post some dyno results with the new turbo soon.

Once at WOT in the higher load cells I think the Mivec settings are pretty close to optimal besides maybe reducing some of the exhaust retard in the JDM maps. Without a good way of measuring changes via dyno or detailed and repeatable datalogging it will be hard to measure changes in the higher load cells at WOT.

Thanks Hiboost for all the tips! So during spoolup, how lean can I safely go? I am logging with a wideband, and was surprised to see high 13's in AFR prior to me richening the map up. I did not get any knock sum during spool though.

I am running the JDM mivec except I zero'd out a lot of the higher rpm/load cells for the exhaust, such that it hits only 0 or -5 cells after hitting peak boost. The car feels a lot more preppy in the low end.

So for now, my main concerns are AFR during spoolup, and some high end knock that seems to randomly occur between 5500 and 7500 rpm's.

you can generally run 12.5-13 upto 10psi, and then quickly drop to 11.3-5 ... then run that flat to redline.

a lot of guys will say just run 11.5-7 - but I find that a little bit of a safety barrier (in case you tuned on a hot/cold day or WB02 was out a bit) is a good thing

Thanks tephra! I richened things up even more, and am starting to wonder if my thinking and logging methodology is flawed. I start my logs at WOT at 2500 rpm's in third gear. This causes me to hit the load cells 100-160 from 2500-3000 rpm's. It runs leaner in this area than 12.5:1. But in the real world, does it really matter? Will I be hitting these load cells normally since I don't usually floor it at 2500 rpm's? I'm usually at 3000 rpm's before I go WOT. Is it dangerous to be leaner such as 13:1 to 13.5:1 between 2500-3000 rpms? I would think the stock fuel map would have the same problem based on how lean it's set at low rpm/load.

Anyone? :) Most of the maps I've seen posted for evo 8/9 keep the fuel pretty lean 14.4 etc at low load/rpm, but I have seen one map that had 12.8/12.9 in the map at 2500/3000 rpm's between 100-120 load. Am I the only one adding a ton of fuel at low rpm/load in order to avoid too lean of an AFR during spoolup? Without the changes, I was seeing up to 13.5:1 at 3000 rpm's with > 10 psi boost.

I won't know AFR values until I get my wideband installed next week but so far after the fuel pump install I can definitely tell my car is running richer at lower RPM's and during spoolup so I'll be adjusting that back to how it was. In general the leaner you can run without causing knock at spoolup will improve response considerably. Adding in fuel just for the sake of adding in fuel to be safe will tend to bog the engine and reduce turbo spoolup.

Thanks! So is it safe to say that I can run lean such as 13.5:1 without worrying too much as long as it doesn't knock? I wasn't getting any knock, but was adding fuel just for the sake of running richer since most say to run 12.5:1 during spoolup :)

wait, what is this tephra logging software I just heard about? where can i get it? what does it log?? this just made me very excited for the day......


"without worrying too much as long as it doesn't knock? I wasn't getting any knock" How are you reading this knock!?!?!?!

You can get tephra's logger here:
https://www.evolutionm.net/forums/ev...x-logging.html

Or, you can buy evoscan for $25 and get external wideband logging support plus other features here:
https://www.evolutionm.net/forums/ev...-load-ipw.html

so whats the general consensus on comparing the afr numbers in ecuflash and what you are reading in Wideband log, are the roughly the same or is it a guessing game again?

It's definitely not the same. You have to use your wideband to get a true AFR, and adjust the map up or down accordingly.

They are NOT the same. You MUST have a wideband to tune the AFR. You cannot realy on the target AFR numbers in the map. Some do this in the Subaru community, but even then the tragets in the map are off from what the wideband reads.

Yeah I think the 14.7 AFR cells are close to accurate at least in the under 100% load cell range, anything else in the 110-320% load cells seems to be quite leaner than what is shown in the tables. My car reads about 11.5 AFR while the table entry is around 10.5 AFR. It would be kind of neat if there was mode you could run that would calibrate it somehow with a few readings at different RPM ranges. For now I just manually adjust the chart to match actual as a rough guide and then print it out as reference for that tune. Saving it that way would blow the engine for sure as it would attempt to run 12.5 AFR's under full boost, not good!