Timing advance would be handy as well, for testing WOT, part throttle and cruise differences to optimize power, spool and efficiency quickly. Of course anyone doing this should have a friend driving while you work the ECU through various changes.

 

oh yeh... all altmaps (and regular versions) are live tuneable

 

haha - the joys of writing M32R ASM at 3:30am...

had to bench my ECU... I think I found the bug and it seems to be working ok for now...

anyone know how to make video of an application??

 

Hello, Live tuning includes alt maps. Awesome.

I'm using "Free Screen To Video V1.2". You can choice the recording format from flv, wmv, wmf, and smf. You also can choice where to record from "entire screen", "window selection", and "selected area". (Though If you want to record high density, it depends on the CPU power.)

 

hrmm so its all working...

except for MIVEC at idle...

when I log the target - i get the right value, when I log actual it just stays the same...

has anyone changed MIVEC at idle (in the maps) and had that change reflected in the logs?

I pushed the TPS so that the revs were at 1250-1500 and my live tuning was ok then...

any ideas?

 

Is oil pressure to low to change the cams?

 

I thought the same thing when I read this.

Is there enough oil pressure (20-30) psi to change the cam positions a significant amount? Also is there a mechanical limit on cam position at idle?

 

thats what I initially thought..

but I dont think so..

The car will generate DTC's for cam advance/retard ABOVE 1188rpm's..

I think the ECU wont allow MIVEC configuration below that RPM...I tried to look at the code, but as usual its complicated and I couldn't really be bothered to get into it...

 

What happens if you flash different MIVEC values in the idle cells?

 

yeah I was going to try that - but didn't.

I think I have tried it before and it doesn't change anything...

 

So its some sort of limit in the code. Interesting. There is something in the Evo 8/9 code that can make MIVEC have a different value than the table during initial startup. I've seen it go to -1.2 deg when it was supposed to be at zero.

 

Interesting stuff!

 

yeh in the code the MIVEC (both exhaust and intake) revert to the Minimum under a certain set of conditions..

initially I just thought it did that if it got an out-of-bounds response from the table lookup but it seems more than that.

when i have some free time I will back trace the code to find out what the condition is, I feel its when the car is "idling", but it could easily be related to just a simple RPM comparison (although I didn't see that when I first looked)

 

ok well I have the basic 3d map stuff working... Its pretty cool, I will take a AppVid tomorrow and post it...

I also bunched all the log items into 1 block of memory, so this means we can log 10 things with 1 call to the ECU (instead of 10 calls)

So hopefully instead of getting 100 items per second (10 log items 10 times a second) - I will get close to 500 items a second (there is a little overhead since CAN packets can only be 12 bytes long, so asking for 10 things (each 2byte) + headers means we need 3 packets...But still, it should give us heaps more data, and CONSISTENT data (meaning all the variables are from the same time instant))

I want to be able to modify the 1d and 2d maps (eg injector scaling, cranking enrichment) as well, but thats just a bit more code...

Its still fugly thou

 

You sleep Evo Ecu code...

Keep up the good work!