RomRaider

The closed loop corrections are done. I saw data points as high as 30% (with an average of 14%) after switching the injectors to the rr recommended 810cc. But most of the sub 2.6V area is now back to where I had it. Idle still needs a tad bit of poking around to cure the -8%.

Now using the wideband, I'm fleshing out the changes to the upper area. But I'm getting weird responses to what I'm doing. Before I went to 810cc I used this same method and was able to get things dialed in quickly. Basically I log MAFgs, WBO2, load, rpm, front AFR sensor (just for fun). I subtract the target AFR (from the proper load cell) from the WBO2, then divide that result by 14.7. This gave me the % I needed to modify that g/s on the MAF scale. This seems to be inching me closer, but I've already added about 14% and I'm still seeing products of my equation telling me to do another 4%. (it always seems to indicate 4% each run through this.)

Am I just digging a bigger hole? Or should I keep going in this direction? The WBO2 still shows when I'm asking for 12.8 the car is giving 13.4.

BTW I'm using WGDC of 0% for now till I get some kind of logical response from my actions..

Since we are comparing ratios, the math is not that hard. Just make an AFR deviation table. Use airboy's spreadsheet to get your target AFR (with the ECU interpolation) for your run. Copy that and put it beside your real logged AFR in the sheet. Real / Target = deviation. Multiply deviation by your MAF G/S. This will get you closer, but remember that the ECU interpolates, so you want a smooth curve.

-Gabe

Why would you divide the result by 14.7?

I meant to divide by the target AFR, not 14.7.. My head isn't on straight this morning

Tried your spreadsheet, pretty amazing. Doesn't work for me because I think I'm missing MAF V.. But I have no idea what data I should be logging and didn't have time to read through the whole thread just yet. Only on page 3. I'm sure its in there.

You only need to log RPM, Engine Load (Direct) and WBO2 to do the AFR interpolation. MAF G/S = RPM*EL/60. Doing it this way gives you enough resolution to see where it is deviating (and by logging EL direct, you get a more accurate representation). Go by the G/S values and look them up in the table. Then make an adjustment row with your MAF scaling pasted in and lookup where deviation is > .01 (IE 1.01, 1.02, .99, .98 etc) and apply the deviation to MAF G/S CEL that is effected. Remember there is lag time between the WBO2 and the MAF, so say you went lean at a reported 4.4K rpm. More than likely that leanness occurred about 100 rpm prior. Takes some time to get it just right, but it works.

Doesnt airboys spreadsheet already do this? On the AFR interpolation sheet it gives you the percentage that you are off by. Apply that to your MAF scaling and you achieve the same result, no?

If you log MAFv as well it does. But you do not need MAFv. The less you log, the more resolution you have...

Wow that makes my head hurt... I get a list of percentages now, but there is still a large gap between WOT and the MAF scaler in RR, unless the spreadsheet can be trusted to come up with a generally acceptable starting point for non-WOT data.

I take it you have to manually apply the percentages and smooth that out right? Did I miss a place to paste my MAF table and have it do all that for me?

Is there some kind of reference manual for that spreadsheet, other than bugging the @#$% out of you guys? I'm pretty dense when it comes to excel, if it were SQL, I'd be in heaven...

If you start out at 2k rpm on your pull, you will have values in the 80+ G/S range. That is what you want to modify by hand. The MAF scaler in RR can do 100 G/S and below (as long as you stay in closed loop).

Yes, you manually apply the percentages to G/S. Some people find it easier to just modify the Voltage instead of G/S. I just like doing it the "factory" way since all of the ROMs I have seen have the same voltage "steps", but differing G/S values.

As far as copying and pasting into excel and from excel, you just copy the MAF table in RR, paste into excel, then move the voltage row 1 cell to the right, and move the "[Table2D]" to the first cell in the voltage row. Then you can paste back into RR after applying your multiplier.

I logged what you said from 2K to about 5k, WOT. Airboy's spreadsheet is refusing to accept the LC-1 data (doesn't show up in any of the windows on the right) no matter what order I do anything in.

As long as you log the wideband and it saves in an excel file then I dont see why it wouldnt work. If you open up a regular log does the LC-1 show up?

This is what csv I'm using, load-direct, rpm, wbo2.

Gabedude, is there some figure ($$) that I could offer you for some direct assistance, beyond what you have already done?

Your LC-1 show up fine for me. I would also try just logging times when you are on the throttle and not just giving it gas and letting off/shifting like this log appears. That should make it easier to sort through the data.

when you select "grab headers" you will need to manually set the LC-1 as your AFR source

maybe I fudged up my spreadsheet somehow. I'll try a fresh copy.