AI: holden vy-l36-060a-enhanced-asm-patches github public

KingAustraliaggGG
Posts: 49
Joined: Fri Nov 15, 2024 2:19 pm
cars: 323i 1999 e46 sedan m52tub25 ms42 hx35w
328ci 1999 e46 coupe m52tub28 ms42
vy 2002 holden wagon executive l67 bottom end/36 top swap corn fed boost manifold ex44 external turbo hx40 holset
vs holden sedan turbo gt3582r

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by KingAustraliaggGG »

antus wrote: Wed Jan 21, 2026 1:05 pm you are posting a lot of stuff but no results. get something to work as a starting point then keep going. your signal to noise ratio is not good. quantity does not equal quality.
## Windows Fix for udis (tranter/udis)

**Original code crashed on Windows:**
```python
# Avoids an error when output piped, e.g. to "less"
signal.signal(signal.SIGPIPE, signal.SIG_DFL)
```

**Error:** `AttributeError: module 'signal' has no attribute 'SIGPIPE'`

**Fix applied:**
```python
# Avoids an error when output piped, e.g. to "less"
# Windows doesn't have SIGPIPE, so wrap in try/except
try:
signal.signal(signal.SIGPIPE, signal.SIG_DFL)
except AttributeError:
pass # Windows doesn't support SIGPIPE
```

**Location:** udis.py line ~43

then seperate to that my tool already worked. so hook point and stuff on the github and all the info is correct unless it says assumption or theoretical.
i made mine with the hc11 documentation. then that tool it says doesnt run on windows 11. just made it run on windows 11 and its same as my tool. good for third party verifications fast


VERIFIED: Min Dwell Init Disassembly Test

*udis output (correct):*
asm
4839 CC 00 A4 ldd #$00A4
483C FD 14 4A std $144A
483F CC 00 24 ldd #$0024
4842 FD 14 48 std $1448

January 21, 2026: my hc11_disassembler_complete.py matches udis exactly:*
asm
; From hc11_disassembler_complete.py output
4839 CC 00 A4 LDD #$00A4
483C FD 14 4A STD $144A
483F CC 00 24 LDD #$0024
4842 FD 14 48 STD $1448
4845 CC 00 04 LDD #$0004
4848 FD 16 F8 STD $16F8
484B BD 2F BC JSR $2FBC
484E 39 RTS

......................
mine uses capitals donno if that matters let me know
User avatar
antus
Site Admin
Posts: 10012
Joined: Sat Feb 28, 2009 10:34 am
cars: TX Gemini 2L Twincam 8psi
TX Gemini SR20 18psi
Datsun 1200 Ute
Subaru Blitzen '06 EZ30 4th gen, 3.0R Spec B
Subaru WRX 2007

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by antus »

ok, maybe thats good. still tl;dr from here but do you have a bin you've tested that you think people will want for some feature to post up for testing?
Have you read the FAQ? For lots of information and links to significant threads see here: http://pcmhacking.net/forums/viewtopic.php?f=7&t=1396
KingAustraliaggGG
Posts: 49
Joined: Fri Nov 15, 2024 2:19 pm
cars: 323i 1999 e46 sedan m52tub25 ms42 hx35w
328ci 1999 e46 coupe m52tub28 ms42
vy 2002 holden wagon executive l67 bottom end/36 top swap corn fed boost manifold ex44 external turbo hx40 holset
vs holden sedan turbo gt3582r

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by KingAustraliaggGG »

antus wrote: Wed Jan 21, 2026 1:27 pm ok, maybe thats good. still tl;dr from here but do you have a bin you've tested that you think people will want for some feature to post up for testing?
anyone got a oscillator and bench setup. i am still validating my python auto patcher and xdf patchlist works before i even think of pushing it. :thumbup:

also any info you see on there about things not really related to the vx/vy $060a enhanced 1.0v, that has errors and you know the right info please correct me on facebook or here and i will fix it fast.
User avatar
Rhysk94
Posts: 235
Joined: Sun Mar 15, 2020 9:25 am
cars: VE SS-V
VY V6 Turbo “Ecojet”
FG XR6 Turbo
VY V6 Stocker Farm Hack
VY V6 Turbo Competition Drift Car
VY V6 L67 Seat Time Drift Car
N80 Toyota Hilux
VZ SS
Location: Perth

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by Rhysk94 »

Few questions.

Why Alpha-N ?
Why does that entire gitub page look like it has been copy & pasted straight from an Ai Generator?
There is no actual files there for anyone to test or any information that wasn't already publically available... which also hence... Ai generated ?

Also you state on there ghost cam & idle spark correction... that is not how it has been done on these pcm's, you might want to remove that incorrect info from there.

Also remove crediting my name for the 6375rpm max, that isn't my finding, I believe I originally got told that by Chrome or The1. They should get the credit for that, they have also stated that on pcmhacking which could be referenced.
Ken
Posts: 238
Joined: Mon Dec 16, 2013 3:05 pm

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by Ken »

Rhysk94 wrote: Thu Jan 22, 2026 7:49 pm Why does that entire gitub page look like it has been copy & pasted straight from an Ai Generator?
It might be because his hobby / business seems to be AI oriented according to a quick Google search.
KingAustraliaggGG
Posts: 49
Joined: Fri Nov 15, 2024 2:19 pm
cars: 323i 1999 e46 sedan m52tub25 ms42 hx35w
328ci 1999 e46 coupe m52tub28 ms42
vy 2002 holden wagon executive l67 bottom end/36 top swap corn fed boost manifold ex44 external turbo hx40 holset
vs holden sedan turbo gt3582r

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by KingAustraliaggGG »

Rhysk94 wrote: Thu Jan 22, 2026 7:49 pm Why Alpha-N ?

Also remove crediting my name for the 6375rpm max, that isn't my finding, I believe I originally got told that by Chrome or The1. They should get the credit for that, they have also stated that on pcmhacking which could be referenced.
alpha n is mafless. can run bigger throttle body or itbs and actually make power instead of no difference at all. can get more power sometimes than a stock maf up high rpm for more hp on dynos with the right hardware and bigger injectors.

fixed it.
charlay86 July 25, 2012, 6375 rpm is the limit of the internal variable in the PCM (255 * 25 = 6375)" Topic 2544
Chr0m3 July 5, 2024, Same explanation, 12 years later in Topic 8567

Fair point for the lumpy idle compared to ghost cam- I've clarified the terminology now:

LUMPY IDLE (your approach):
- XDF parameters only, no ASM patching
- Expose KSARPMHI/KSARPMLO multipliers, modify in TunerPro
- Result: Slow ~1Hz "lope every second"

GHOST CAM (my research goal):
- ASM patch adding BMW-style RPM-delta lookup table
- Non-linear response with 45°+ swing
- Goal: Fast aggressive cam sound like LS/BMW ghost cam

Your lumpy idle works - you proved that. I'm researching whether true
ghost cam (fast lope) is possible via ASM patch on these PCMs.

The XDF parameters you exposed are already in v2.09a $060A:
- KSARPMHI @ 0x6525 (stock: 0.04 °/RPM)
- KSARPMLO @ 0x6527 (stock: 0.04 °/RPM)
- KSAOLDB @ 0x6529 (stock: 512 RPM deadband)
- KSCORLIM @ 0x652B (stock: 5.27° limit)

Stock values are neutered - that's why the response is slow even with
your XDF changes. The underlying algorithm is linear P-gain, not a
lookup table.

i use ai cause i dont want to be spending the next 10 years on this
User avatar
pman92
Posts: 667
Joined: Thu May 03, 2012 12:50 pm
Location: Castlemaine, Vic

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by pman92 »

KingAustraliaggGG wrote: Mon Jan 26, 2026 8:27 am i use ai cause i dont want to be spending the next 10 years on this
Lol :wtf:
How much time do you think AI has saved so far?
Like what percentage of these patches has actually been tested or at least looked at in ghidra or similar by a human to verify it's actually doing what it is meant to?

Or are we just blindly swinging the hammer and pretending we're getting the job done, but not actually looking to make sure we're hitting the nails?

I'm not trying to have a dig, but when everything sounds like it's been regurgitated out of chatGPT it is impossible to take seriously.

Maybe you have verified some of whats here, and if so great, but I haven't seen any evidence of that yet?
If not, can we at least keep it contained to this topic. As I said last time (and this is just my 2 cents), this old school forum is one of the last places that isn't an AI echo chamber, and I would hate for all the old wealth of information posts to be buried under a ton of slop.
KingAustraliaggGG
Posts: 49
Joined: Fri Nov 15, 2024 2:19 pm
cars: 323i 1999 e46 sedan m52tub25 ms42 hx35w
328ci 1999 e46 coupe m52tub28 ms42
vy 2002 holden wagon executive l67 bottom end/36 top swap corn fed boost manifold ex44 external turbo hx40 holset
vs holden sedan turbo gt3582r

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by KingAustraliaggGG »

When did GitHub go up, 10 days ago. Already found stuff no one has put public and put it public ai or not.
User avatar
Rhysk94
Posts: 235
Joined: Sun Mar 15, 2020 9:25 am
cars: VE SS-V
VY V6 Turbo “Ecojet”
FG XR6 Turbo
VY V6 Stocker Farm Hack
VY V6 Turbo Competition Drift Car
VY V6 L67 Seat Time Drift Car
N80 Toyota Hilux
VZ SS
Location: Perth

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by Rhysk94 »

8-)
KingAustraliaggGG wrote: Mon Jan 26, 2026 8:27 am
Rhysk94 wrote: Thu Jan 22, 2026 7:49 pm Why Alpha-N ?

Also remove crediting my name for the 6375rpm max, that isn't my finding, I believe I originally got told that by Chrome or The1. They should get the credit for that, they have also stated that on pcmhacking which could be referenced.
alpha n is mafless. can run bigger throttle body or itbs and actually make power instead of no difference at all. can get more power sometimes than a stock maf up high rpm for more hp on dynos with the right hardware and bigger injectors.

fixed it.
charlay86 July 25, 2012, 6375 rpm is the limit of the internal variable in the PCM (255 * 25 = 6375)" Topic 2544
Chr0m3 July 5, 2024, Same explanation, 12 years later in Topic 8567

Fair point for the lumpy idle compared to ghost cam- I've clarified the terminology now:

LUMPY IDLE (your approach):
- XDF parameters only, no ASM patching
- Expose KSARPMHI/KSARPMLO multipliers, modify in TunerPro
- Result: Slow ~1Hz "lope every second"

GHOST CAM (my research goal):
- ASM patch adding BMW-style RPM-delta lookup table
- Non-linear response with 45°+ swing
- Goal: Fast aggressive cam sound like LS/BMW ghost cam

Your lumpy idle works - you proved that. I'm researching whether true
ghost cam (fast lope) is possible via ASM patch on these PCMs.

The XDF parameters you exposed are already in v2.09a $060A:
- KSARPMHI @ 0x6525 (stock: 0.04 °/RPM)
- KSARPMLO @ 0x6527 (stock: 0.04 °/RPM)
- KSAOLDB @ 0x6529 (stock: 512 RPM deadband)
- KSCORLIM @ 0x652B (stock: 5.27° limit)

Stock values are neutered - that's why the response is slow even with
your XDF changes. The underlying algorithm is linear P-gain, not a
lookup table.

i use ai cause i dont want to be spending the next 10 years on this
Nope, still incorrect for ghost cam. Your alpha n reasoning is still incorrect too
KingAustraliaggGG
Posts: 49
Joined: Fri Nov 15, 2024 2:19 pm
cars: 323i 1999 e46 sedan m52tub25 ms42 hx35w
328ci 1999 e46 coupe m52tub28 ms42
vy 2002 holden wagon executive l67 bottom end/36 top swap corn fed boost manifold ex44 external turbo hx40 holset
vs holden sedan turbo gt3582r

Re: holden vy-l36-060a-enhanced-asm-patches github public

Post by KingAustraliaggGG »

When you say “ghost cam incorrect”, do you mean I’m using the term wrong, or do you mean the mechanism won’t work on these Delcos? I agree the existing method people are using on 060A is basically idle spark correction driven off RPM error, your style of approach, and I’m not disputing that. What I’m talking about is an optional ASM experiment to reshape that correction law using an RPM-delta lookup table, so the response can be more nonlinear or asymmetric than the stock proportional behaviour. If you think that’s not viable, which part specifically is wrong, the hook location, the spark unit scaling, or the control approach itself?

On Alpha-N, I’m not claiming it magically adds power. It’s just a different load estimation strategy that can make sense when MAF or MAP becomes unreliable or hits limits. Bigger injectors alone aren’t a reason to go Alpha-N either, they can work fine with MAF if the injector data is set up correctly and the MAF isn’t saturated. The only point I’m making is that some engines have made strong NA gains on Alpha-N setups in the real world, and I haven’t seen a proper public Ecotec/060A Alpha-N effort that’s been validated with logs and dyno runs, so on this platform we won’t know either way until someone tests it properly.

can you explain it more than just its wrong.