I could give it a shot and see...I dunno, for some reason I think my problem goes a little further than adjusting the fuel map...my next approach is to try a totally stock xml and see if that works (one without all the patches etc) Currently I'm running v1.29a with: tephra v5.10, JDM map sensor patch, leanspool disabled via periphery bits along with EGR stuff for the map sensor, MUT table defined and setup for the 2byte stuff.

 

I'm wondering why AFRMAP is not tracking the fuel map. At 4000 rpm, 318 load, AFRMAP is at 11.5, but there are no cells in the fuel map close to 11.5 unless you go down to 2500 rpm. I had thought AFRMAP was the value taken directly from the fuel map, but possibly it is the value after other multipliers are included?

 

This is an aside, but are you seeing 1 degree pulled for 1 knock count or is your timing map giving you the 4 degree value in the log?

 

Nope your right, 1 degree of knock for me pulls 1 degree of timing...which I found weird because I was told that it takes 3 counts of knock to pull a degree.
I had asked this question the other day and this is what I got:
https://www.evolutionm.net/forums/ec...ls-timing.html

 

The reply you received in the knock post by John and then me was the correct answer. Did you read John's knock routine disassembly thread for more info?

 

I did read the post, it was very informative...since i'm still running stock turbo/cams/internals I decided not to look too much into the knock tables since chances are whatever knock I do register is more than likely not from harmonics but real.
Do you think accel enrichment would have anything to do with my problem? I remember reading somewhere in another post something related.

 

You know, I have logged a 9 and saw what appeared to be a similar relationship (1 degree per knock count). I just chalked it up to a slow logging rate.

Now I don't want to sound like I am disputing John's findings, because I am not, but looking at the log above it really looks as if 1 knock count equates to 1 degree. It is also seems unusual to me compared to all the logs I have done over time, in that knock counts of 1 normally don't last so long before decaying to zero.

 

How can you come to that conclusion without even seeing his timing map? There are spots in his log where one might think that 1 count pulls nothing, 1 count pulls 2 degrees, or 1 count pulls 1 degree. But, you just can't tell without knowing what the timing is supposed to be. You have to remember that the ECU is interpolating between many cells. You're rarely smack dab in the middle of a cell. It can be interpolating from 4 nearby cells. And going through the map that quickly, you can see where the true timing is difficult to get to. Throw in the various loads and some people logging the wrong ones, etc, and there simply isn't a definitive way to state simply by looking at a log, how much timing 1 count of knock pulls. That's where the disassembly of the ROM comes handy. That way we know for 100% sure what is going on.

I think it all just comes down to several factors such as our logging speeds and resolutions (for example, logging timing is 1 degree increments whereas the ECU is working with fractions of a degree). Simply looking at a log though without at least tracing the path through the timing map tells you nothing. First you need to know what the timing should have been, then look at what it actually is. It's OK for extreme cases where you are actually right in the middle of a cell and certain that you are logging the right load and know all of the corrections being applied (coolant temp, air temp, etc) and have like 15 counts of knock and compare it to your map and you are exactly 5 degrees off. But, the 1 count of knock thing and everything I mentioned just is too difficult to determine from one of our logs.

 

check your voltage! looks like your battery correction offset will be messing with your AFR's if it isn't tuned properly.

get your power supply figured out and then try again.

 

I am not coming to a conclusion, or stating definitvely, just making an observation. What I noticed first is the fact his timing is dropping with load dropping/rpm rising. Once you pass peak load you are always able to add timing due to rpm increasing and load reducing. Keeping this in mind and looking at his timing it suggests that timing is being pulled, but it also could be that the map needs more work. Then I noticed that the 1 knock count lasts what seems like quite a long time, which is also unusual from my experience.

Your point is valid though, that there are many things at play so it is not possible to conclude that i count is equating to 1 degree in this case.

I am actually more interested in why his AFRMAP does not correspond to his fuel map. I didn't think AFRMAP was subject to additional corrections, but I would like to know if that actually is or is not the case.

 

Hope you figure it out raasfaas, I'm dealing with the same issue. I use to think it was because of my meth activating but I'm still getting an awful dip at peak load. I determined it isn't the fuel map, because no matter what rpm I start at it dips very rich once it hits peak load.

 

I'm not sure what offset your referring to, could you point me to it. I was just looking at the battery level that evoscan is logging.

l2r99gst, I get what you're saying about the relation between knock and timing, and there are instances where I get knock and it doesn't pull any timing and some where it pulls more than 1 degree.

wreckleford, I would post my timing map but this log is a couple weeks old and I have made changes to the map since so tracing through it wouldn't reflect the right #'s I'm not sure on the AFRMAP issue, I hope someone more experienced can clarify.
Like I said, at first I thought it was meth related but it happens also when I'm on pump gas, its just not as pronounced.

 

I'm not sure if the MUT value we log is final or initial, but I imagine it would be final including lean spool, etc etc.

 

You are refering to AFRMAP right?

 

Wreckleford, on the AFRMAP issue, since you mentioned it I just went back through my evo9base's XML and realized that the endian for "AFR" (among a few other items) in the scaling list is set to big and I'm on v1.29a which means it should be little...do you think this could be the cause of my AFRMAP not correlating with my actual fuel map?