I have experience with the 8051 and was actually hoping you would select this route. I would like to discuss your c program with you when we get a chance. I assume we would use your developed sensor input/output scheme from your experience with the c compiler version, I really would not have the time or knowledge to learn what needs modified and when. I assume you have a grasp on this. How many input/output sensors will be attached to the system in the final operating system? Do you have a vision of your operator interface or will this be a black box? The software does not bother me, the noise picked up by various transducers does.

I live near Youngstown, Ohio. Don't know if this hurts or helps... Finally, I just purchased my S2000 the end of June. I do not have much experience with the s2ki.com forums. These are my first posts (Just been reading for a couple months...learned a lot so far...watched and Evo autocross school one day...).

 

A buddy of mine saw this thread and offered to take a whack at it. If he decides he doesn't have the time or energy, I'll get back with you guys.

FWIW, I don't have any interest in having a user interface on the box. All the settings would be adjusted through a windows based application connected to the black box through a serial interface.

 

Just out of curiosity is wideband the only input, ordoes it do any pre-emptive tuning based on MAP signal or throttle position?

I guess the MAP signal would make it more attractive to SC and turbo owners.

 

It is essentially a V-AFC with a WBO2 input and much finer resolution of MAP correction at every 100 RPMs. Basically, it is overkill. Here is an output of my WBO2 log from about 3K to redlne (on the street). I intentionally set a stair step in target AFR to see how quickly and smoothly it would make the transition. I didn't expect it to do it this well.

 

the curve looks smoother because the X-axis on the plot is stretched out

the VAFC controls VTEC switch which no one mentioned that gives you more power at earlier rpm

berz; what do you mean a road with constant load. u r better off street tuning because you can get real conditions. Try it on uphill, flat road, with your fat booty call who slurps mean wang in the shotgun seat, etc. On dyno its tuned to one scenario, which is fine as it also gets the job done. if you have wide band lamda, you can monitor and retune at any time if you encounter unfavorable conditions( extreme heat, elevation changes, barometric fluctuations)

you could add a Comptech ESM to try to add more fuel than just a few percent, though ECU still wigs out at extreme settings on Comptech SC cars.

 

Looks to me like gernby's controller does this, too. The dyno graph above looks like the VTEC crossover is set at ~5700. If the controller doesn't do VTEC switch, I'm sure it would be relatively easy to add that functionality.

 

Which X axis is stretched out? The variation in AFR is only about +/- 0.25 points. The variation on a stock curve is about 2 full points.

I'm not sure what you meant about the VTEC engagement, or if it was in reference to my controller, but my controller does do VTEC engagement as well.

 

Now I see what you are talking about regarding the X-axis being stretched, but notice also how stretched the Y-axis is. The AFR fluctuation is exhagerated by the scale. The 1st plot shows an AFR range of 10-20, while the 2nd shows about 12-16.

 

I assumed he was talking about the Y-axis being stretched. Since it has a larger range, it makes the AF curve look smoother. Still, even with the wider range you can tell that the AF ratio didn't value more than 1/4 point from your "preferred value". That's very impressive considering that the VAFC varied as much as a full point (specifically around 6000-6500 RPM).

 

Think about it ... stretching the X-axis does make a horizontal wave look flatter. It is true that the 2nd plot is stretched some. However, stretching the Y-axis makes a horizontal wave look LESS flat. If the X-axis is stretched by the same amount as the Y-axis, the wave looks unchanged. In my 2nd plot, the Y-axis is stretched more than the X-axis, so it really looks LESS flat than it would if it wasn't stretched at all.

Basically, the 1st plot is a scanned in printout, and the 2nd plot is a screen print. What I need to do is put both plots on the same screen shot, but I can't find the Dynojet run viewer on DynoJet's web site.

MXT_77, you have my dyno run files, right? Would you put those 2 pulls on the same plot and post it?