Yes, I checked about 4-5 points in his log and this seems to hold true.

It's harder to tell in places where there is 1 count of knock, since EvoScan will only report timing in whole numbers, but it does look like the Evo ECU is acting as I suspected.

I will have to do a bunch of logs myself this weekend.

I haven't been on aktivematrix....what is the latest with getting map tracing, once we verify this load forumla a few more times and get the constant nailed down exactly, if not already?

Eric

 

I've been disassembling the serial routines this afternoon and have some things to work on. If we can hook into this we can find code that references bits of map we're interested in and chuck out bits of data I reckon - custom request IDs for Evoscan would probably work well. I'd halve the airflow Hz as well to start with!

 

so that .xml file will work for my USDM ECU??

I noticed you added AFRMAP, SPEED, & MAP, but didn't see any of the wideband support stuff. Anything else you "tweaked"?

 

He also tweaked the Octane number.. scaled it to max out at 100 instead of 255. Either way is fine with me, but I wanted to mention it so nobody freaked out when they saw "100" for their octane number.

 

FYI, I derived the formula today for g/rev . Here it is.

g/rev = (AFRMAP * (IPW) * (Inj. Scale))/ 20408.2

AFRMAP is in a/f points (i.e. 12:1)
IPW is in ms
Injector Scale = constant (stock USDM 516 cc)

This formula assumes density of gasoline to be .735 g/cc

This may or maynot be useful. Someone may have already derived it. But all this talk got me thinking and I decided to write it down.

Brian

 

Sorry, Always in Boost, I removed the stuff I didn't need or unticked it. You can copy and paste what you want from my xml file into yours and that might work better.

Thanks for that Brian, I was wondering about working something like that out to get more precise numbers.

 

jcsbanks - i tried out your new AFRMAP calculated load on a more rigurous level last night and noted I am getting computed values in the 340-375+ load cells at WOT operations!

Maybe my Excel formulas are off?

=ABS(PRODUCT(2550,G2-0.4)*1/P2)

 

My injector scaling number is 542 (JDM/UK). Maybe it needs adjustment for different ECUs? Can you post the log?

The example shown earlier by Rob had some fairly flat timing, but seemed to line up OK.

 

Log enclosed, unfortunately I cannot post my maps (vendor tuned).

 

The formula still looks reasonable. The problem in that log is that there is about 1/3 of a second between IPW and AFRMAP being logged, so on fast throttle movements either way the AFRMAP value should probably be paired with the IPW on the NEXT line. You could increase the sample rate in Evoscan, or move IPW and AFRMAP adjacent to each other in the xml file for logging. Evoscan is logging sequentially, but we present the logs on one line with the same time stamp when this is not the case.

 

Okay, I am missing something. I don't get AFRMAP, LOAD, SPEED, OCTANE or CalcLOAD. How did you guys get it, some help please.

 

Oh man, I forgot to post the data.xml file that I updated with all of the new found data items.. the calc load is exactly that, calculated, I'm rewriting the Evoscan log analyzer to do that for you..

 

Thanks Man, Are they going to be posted any time soon or in couple of days.

 

Thanks for the log ST!

 

lol...no probs...beers still on me if you're ever down here.

jcsbank - i will try to redo the file order as well as increase the baud rate and see if this helps.