NefMoto

original thread regarding avoiding ps_w limit:

http://nefariousmotorsports.com/forum/index.php?topic=2747.0

Hit some strange issues on a car. No clue why.

Here is a log of a whole bunch of different revisions of the same file, all using different KFKHFM scalings at the top of the range.

Note the area where ps_w, rl_w, and rl all get locked to the same value.

anybody have any ideas?

(http://nyet.org/cars/images/ps_w-workaround.png)

I'm guessing the problem lies here somewhere...

(http://nyet.org/cars/images/bgmszs-dabgl.png)

My current guess is that DSLOFS=0 is bad, because there are cases where pvdksf_w can actually hit PVDKMX.. and set B_edsl

I still can't make head or tail out of E_lm or E_dk from the FR though... but i don't think they are set because RLNOT is very very low.

which means i need to log

B_neglgrd
E_lkvdk
B_mxrlroh
B_mszsdkb
rlflmroh_w
rldkofk_w
rlfdkroh_w

to see what is interfering with rlfgroh_w

I have nothing to contribute (other than I thought the 5120 was supposed to make this method obsolete?), but I'm posting anyway so you don't have to talk to yourself. Carry on, I'm with you.

(lol)

That looks strange, have you verified axes isnt swapped on KFKHFM?

If I were you I would just return those to stock and scale down KISRM by the amount needed and KFURL/KFPRG up by the same amount, sane MAF readings, sane load and no 2map transition to calibrate :)

Just came to mind, you might be triggering b_neglgrd when scaling down ps_w that way?

CWBGMSZS -> 0

Looks like that inhibits B_neglgrd?

Thanks for the hint, I will try it out (and hopefully analyze the FR so I understand what it is doing).

From looking at bgmszs-neg.lgrd, wouldn't it make more sense to tweak PSSOLNGRD (and/or log grdpssol_w)?

Or is that just to test if B_neglgrd is being set w/o logging it?

I'm not sure this is it; plsol and pssol are not moving around fast enough to generate a big enough grdpssol_w in that region, unless I'm missing something

(http://nyet.org/cars/images/neg-lgrd.png)

it also inhibits whole b_inisabgl

Ah. I see. Looks like I have a lot more reading to do. I'd rather not just zero that w/o knowing what it is doing.

Thanks...

0x10D69 looks like, for future reference.

Looks like bit 1 of CWBGMSZS (the only one set in mbox) enables/disables B_fkmsdks, but not E_lkvdk/B_inilabgl

(http://nyet.org/cars/images/CWBGMSZS-bit-1.png)

I think i'm overthinking things.

Most likely a KFMLDMX -> B_ehfm -> B_hfm issue.

Will recalibrate KFMLDMX more carefully next time i see this.