bn880
09-23-2006, 09:02 AM
Original thread seems to have been lost in the attack, reposting)
Hi again, haven't been around lately as my car was being "refreshed" body wise.
My question is fairly simple... since I run in excess of 30psi of boost at times, and have a 3Bar map under a custom Blueberry cal. I still hit overboost cutout at ~29psi. Are there any specific instructions/ideas on how to totally remove the cutout in Blueberry? I did set MPSHOF to FFh, same prob.
Thanks
Does it shutdown immediately or after some time?
Immediate shutdown when it hits a spike.
Looking at this:
; a is preloaded with baro. Add shutoff to baro,
; if it's more than ff cap it at ff. Then compare
; a with MAP, and if it is above, no overboost
L1665: adda MPSHOF_OverboostShutoff ; (data is ff)
bcc L1669 ; branch if not above ff
lda #$ff ; otherwise, cap at ff
L1669: cmpa MAP ; compare a with memory contents
bhi L1670 ; branch if higher (no overboost)
; otherwise, POSSIBLY overboosting:
adda SHFDLT ; (data is 1a)
bcc L1672 ; branch if greater or equal
lda #$ff ; load a with value -11111111-
L1672: cmpa MAP ; compare a with memory contents
bcs L1668 ; branch if lower
ldb Timer_FuelShutoff ; load b with memory contents
beq L1668 ; branch if equal (zero)
decb stb Timer_FuelShutoff ; store b into memory
bra L1664 ; branch
If I edit the instruction
bhi L1670 ; branch if higher (no overboost)
to just branch, it should skip the overboost routine, but I'm hoping to do this in hex. Recompiling would involve rescaling just about everything.
EDIT: ahh Blueberry comes with an lst output, should be easy enough to locate the instructio in the bin.
Think I got it, changed
bhi L1670 ; branch if higher (no overboost)
to
bra L1670 ; branch (no overboost)
2222 in bin to 2022 (in my bin it was at address F048h - C000h = 3048h)
Will report on how it performs.
Yep, that works too. I had the brute-force bypass for a while too. Even added a TBL entry for the op code. I've been trying to unravel the OB stuff the hard way. I got it to where it will work for a while, but will then shutdown. It is either the 9sec timer (which I think I have ruled out), or it is a limp-in branch caused by a knock detect. I think it is the latter.
Yeah, there are about 3 branches to throw the OB code, meaning all that code before them probably is a cutout, try searching for L1673 (throw overboost code) That is if you use BB code at all.
Good to know the brute force works, I'm now more confident I will get to work without swapping chips. (since I run full boost and no solenoid control, brute force is enough for me)
Report: That did the trick. Back to fuel mixtures. :thumb:
Sweet. Yeah I'd like to still have shutdown if there is something wrong, like I am in MAP limp-in or something. In any case this is better that the hack I had before where the entire engine shutdown routine was bypassed, which disabled the rev limiter.
True, and I understand your desire for limp in mode cutout etc. I mean I'd be going the hard way too if I ran under 29psi all the time.
Hi again, haven't been around lately as my car was being "refreshed" body wise.
My question is fairly simple... since I run in excess of 30psi of boost at times, and have a 3Bar map under a custom Blueberry cal. I still hit overboost cutout at ~29psi. Are there any specific instructions/ideas on how to totally remove the cutout in Blueberry? I did set MPSHOF to FFh, same prob.
Thanks
Does it shutdown immediately or after some time?
Immediate shutdown when it hits a spike.
Looking at this:
; a is preloaded with baro. Add shutoff to baro,
; if it's more than ff cap it at ff. Then compare
; a with MAP, and if it is above, no overboost
L1665: adda MPSHOF_OverboostShutoff ; (data is ff)
bcc L1669 ; branch if not above ff
lda #$ff ; otherwise, cap at ff
L1669: cmpa MAP ; compare a with memory contents
bhi L1670 ; branch if higher (no overboost)
; otherwise, POSSIBLY overboosting:
adda SHFDLT ; (data is 1a)
bcc L1672 ; branch if greater or equal
lda #$ff ; load a with value -11111111-
L1672: cmpa MAP ; compare a with memory contents
bcs L1668 ; branch if lower
ldb Timer_FuelShutoff ; load b with memory contents
beq L1668 ; branch if equal (zero)
decb stb Timer_FuelShutoff ; store b into memory
bra L1664 ; branch
If I edit the instruction
bhi L1670 ; branch if higher (no overboost)
to just branch, it should skip the overboost routine, but I'm hoping to do this in hex. Recompiling would involve rescaling just about everything.
EDIT: ahh Blueberry comes with an lst output, should be easy enough to locate the instructio in the bin.
Think I got it, changed
bhi L1670 ; branch if higher (no overboost)
to
bra L1670 ; branch (no overboost)
2222 in bin to 2022 (in my bin it was at address F048h - C000h = 3048h)
Will report on how it performs.
Yep, that works too. I had the brute-force bypass for a while too. Even added a TBL entry for the op code. I've been trying to unravel the OB stuff the hard way. I got it to where it will work for a while, but will then shutdown. It is either the 9sec timer (which I think I have ruled out), or it is a limp-in branch caused by a knock detect. I think it is the latter.
Yeah, there are about 3 branches to throw the OB code, meaning all that code before them probably is a cutout, try searching for L1673 (throw overboost code) That is if you use BB code at all.
Good to know the brute force works, I'm now more confident I will get to work without swapping chips. (since I run full boost and no solenoid control, brute force is enough for me)
Report: That did the trick. Back to fuel mixtures. :thumb:
Sweet. Yeah I'd like to still have shutdown if there is something wrong, like I am in MAP limp-in or something. In any case this is better that the hack I had before where the entire engine shutdown routine was bypassed, which disabled the rev limiter.
True, and I understand your desire for limp in mode cutout etc. I mean I'd be going the hard way too if I ran under 29psi all the time.