RomRaider

I posted this on the AZ1G202G thread, but thought I'd give it wider exposure, and potentially help others track down this parameter in their roms.

For all the wrong reasons, after some testing, I found the Knock Sum Per Cylinder in my Rom - in the later roms, this is not a single value, but one per cylinder (so 4 seperate parameters for the H4's). I did some fairly extensive logging, and I believe this is also the correct sequence, as cylinder's 4 and 2 seemed to pick up the most knock sum.

To find it is fairly straighforward - go to the location of the Fine Correction Load Column. Follow the Xref back from the first value. Above it (at least in the H4's) you will see a sequence of 4 Ram parameter addresses with an Xref to the first parameter - these are the knock sum parameters:



Definitions for AZ1G202G:

Could someone be so kind as to find the addresses for knock sum in my rom AZ1G400W?

I’m IDA illiterate.

Thanks

Dave.

ROM:00059228 off_59228: .data.l unk_FFFF7C1D ; DATA XREF: sub_3C7F4+120o Cylinder_1
ROM:0005922C .data.l unk_FFFF7C1F ; Cylinder_3
ROM:00059230 .data.l unk_FFFF7C1E ; Cylinder_2
ROM:00059234 .data.l unk_FFFF7C20 ; Cylinder_4

I noticed you have the knock sum referenced as cyls. 1-4 as opposed to A-D used with per cylinder injector offset and ignition timing. If knock sum does indeed indicate each cylinder#, could the tree be traced to verify which # corresponds in the alphabetical sequence? It would clarify the age old debate of 'firing order' vs 'cyl. bank'.

Thanks Heaps td-d,

I haven't tried it yet but I've made my logger deff’s so I hope they work.

I was so confused by my FBKC values last time i tried to adjust my timing i gave up in the end.

I pulled a lot of timing out of my map and still had loads of FBKC I now know it was probably due to pre-emptive FBKC.

Thanks Heaps for all your good work I'll very interested in any further findings.

Dave

I guess A-D would probably have been more accurate - but what makes me think it might be 1 to 4 is that on my setup, 2 and 4 showed the highest counts, which would make sense.

To only way to verify would be to trace the logic back and also log to verify - I'm carless at the mo...

Any luck getting these for an AE5I910V? ...I'll get IDA one of these days

Here you go, in red:

ROM:0006062C .data.l unk_FFFF7D8D ; Cyl_1
ROM:00060630 .data.l unk_FFFF7D8F ; Cyl_3
ROM:00060634 .data.l unk_FFFF7D8E ; Cyl_2
ROM:00060638 .data.l unk_FFFF7D90 ; Cyl_4

Thanks. I will add these to my logs when I get back.

I meant to give some feedback on this, but slipped my mind - I'm pretty certain it's cylinder order - since the 4 parameters are lined up at an offset in the following order:

1-3-2-4

e.g. :


Sound familiar?

Hi Oren,

can u please assist in pulling these out for AZ1J500J?

Thanks.

Thanks Oren.

Hi,

Seems most likely only applicable on 32bit canbus vehicle?
Could it be applicable on A8DH200Y and if yes how is the address?

Thanks,
Sega

The earlier roms only have a single knock sum value, not per cylinder.