yes i do actually. it's part of my job lol.

http://forums.clubrsx.com/showthread.php?t=648077

go by a/f corrected.

with regards to knock, i'd just pull timing from the actual timing map not from the individual cylinder trims. i've toyed with the individual cylinder trims on the dyno and there is some power to be had. but first dial in the main ignition timing tables first. then go back and start playing with the cylinder trims.

as far as increasing timing at full throttle, yes columns 9 and 10. sometimes 8 too if you're at higher elevation.

 

Brian, have you verified the AF Corr value in FlashPro? I can't say for other S2000's, but AF Corr is NOT correct on my car. The most obvious indicator of AF Corr is wonrg is that I can make a 10% increase in fuel, but AF Corr will only show a small change, but AF will show about a 1 point change. I also installed my aftermarket WBO2 (Innovate LM-1), and it reads MUCH closer to AF than AF Corr.

 

i cannot comment with regards to the s2k and flashpro because i have not used a flashpro with a s2k. with the 06+ civic si's the a/f and a/f corrected tend to read about the same around the 12.8-13.0 mark which is what i target for naturally aspirated cars.

on a car i worked on recently with flashpro i was going by the dyno's wideband and i was targeting 13.0. i was looking at my logs and the logs on the flashpro recorded 12.8 across the boards. then again the sniffer wideband will read a hair leaner. so 13.0 sounds about right.

now with forced induction. i go by a/f corrected all the time. the a/f on flashpro will read 10.5 and the a/f corrected will read about 10 percent leaner and is usually close to what the dyno's wideband reads or so i've seen.

 

I don't mean any offense, but I don't think it's wise to assume that the AF is equally incorrect on the S2000 as it is on the Civic. I'm pretty sure the sensors are actually different too.

When I first started using the FlashPro, I datalogged my car with only an intake. The AF Corr value showed that my stock tune was giving AF Corr values that would go up near stoich between 6-7K! I datalogged my brother's 100% stock '06, and it did the same. In fact, my brother's was 13.65 at the richest point (8K+) and 14.8 at the leanest point (6100) Then I noticed that if I used the AF Corr value in FlashPro Manager for fuel suggestions, it's recommended fuel changes were always about half what they should have been.

For example, if my target value for AF Corr was 13.2, and my datalog showed AF Corr was 13.6, then it would suggest that I make a +4% change (IIRC). I would apply the change, but the AF Corr value would only go to 13.4, then it would suggest another +2%, which I would apply. But then AF Corr would only change to 13.3, so it would suggest another +1%. By the time I actually got it down to my target, I had to add about twice what FPM suggested at the beginning.

However, if I tell FlashPro Manager to use AF for fuel suggestions, they are almost perfect. If FPM suggests +5%, and I apply it, it will be almost right at the target. That said, I do set my target richer than I would if I KNEW that AF was correct. I have been tuning for AF = 12.5, which gives an AF Corr value of 13.6.

Here is a graph showing the link between AF and AF Corr.

 

I found exactly the same thing on my car, when I first got my flashpro I did some datalogging and found that during a 3rd gear pull my AFR Corr values would be almost 14:1 where as my AFR values were as you would expect, around 12.8:1-12.9:1.

 

[QUOTE=gernby,Oct 28 2010, 12:03 PM] I don't mean any offense, but I don't think it's wise to assume that the AF is equally incorrect on the S2000 as it is on the Civic.

 

I realize we've gone totally off topic from ignition tuning to fuel tuning, but since this is such an old thread ... I guess it doesn't matter too much.

I've found that if you set up the target throttle map so that you can easily / repeatably run straight up the load lines over and over again while datalogging on the street (where there's no traffic or stop signs), then you can get very good fuel suggestions from FlashPro Manager's lambda overlay.



When I used this throttle map, I could go straight to a load line, and hold it there as the RPMs climbed. The resulting fuel suggestions wound up being very good. After applying the suggestions from just 1 datalog (on the left below), my very next datalog showed a huge improvement (on the right below). Most of the time was spent applying the suggestions, so it would be VERY nice if Hondata would add a feature to the lambda overlay like "Apply suggestions to selected cells".