OBD 1.5 Tuning

[caffeine]

Also an rpm activated switch would be nice in nAst1 . Could be used for meth or other things

[mrogersmike]

Hmmmm.... I'm with caffeine. I have 2 turbos here and a 2 bar map. I'm asking as well. Can I help in some way?
I can get a maf easy enough as well. I have some bins from the dsm days. I'll look at them so I know what's needed too. I'm sure you guys already do.
I don't know alot, but i'm willing to learn/help if I can.

[caffeine]

If you set me up with where in the assembly code to make the changes and a disassembly id be curious to attempt it myself. I have a basic knowledge of x86 assembler already.

[robertisaar]

I'll ask for it if you're willing

which one?

Also an rpm activated switch would be nice in nAst1 . Could be used for meth or other things

something i've had on the backburner for a while..... i was planning on setting up a couple of general purpose 1/2/3 condition values with user selectable greater than/less than/etc choices, but i guess i haven't gotten around to it yet?

Hmmmm.... I'm with caffeine. I have 2 turbos here and a 2 bar map. I'm asking as well. Can I help in some way?
I can get a maf easy enough as well. I have some bins from the dsm days. I'll look at them so I know what's needed too. I'm sure you guys already do.
I don't know alot, but i'm willing to learn/help if I can.

i could always use alpha/beta testers. preferably, i would like someone with a good N/A tune to change only to a 2BAR MAP and make the calibration match it to make sure it will act as intended rather than risk something like boost occuring but no additional fuel added, etc.

If you set me up with where in the assembly code to make the changes and a disassembly id be curious to attempt it myself. I have a basic knowledge of x86 assembler already.

with the P66 V6, the only thing that bothers me is that a tiny mistake can mean having to desolder the PROMs and throw them into a discrete burner. with a MEMCAL based PCM, the worst that happens is the code doesn't work correctly and usually get thrown into limp-home mode. revise code, try again.

i've been VERY careful and selective about what i've patched into the P66 V6 code thus far, but you never know when you'll make the mistake that causes the processor to freak out. i don't see any kind of backup mode with these PCMs other than perhaps the illegal instruction/COP timeout getting asserted and making the code try again. if that keeps happening..... good chance that the bootloader used to upload the calibration would never get in to do its job.


if you have the skill, burner and some time..... different story.

[mrogersmike]

I will do my best to get my tune done by the end of next week with some logs so there is something to start with.

[caffeine]

I don't currently have an engine that runs on a P66 PCM but I do have the PCM and I have a chip burner that burns the 27SF512 chips and others so I don't know if that burner would suffice.

I've kinda been wanting for a while to try setting up an engine test stand to run an engine out of a vehicle. Would be a cool way to test changes without putting a DD in danger.

[mrogersmike]

My rodeo is just a toy for me so I have no problems testing whatever is needed.
Maybe you gents can educate me on my understanding of what the PCM needs for boost. I'm somewhat confused. A map sensor to my knowledge produces a signal between .02vdc-4.94vdc. This signal range does not change when switching to a 2 or 3 bar map. With this in mind, the PCM is seeing the same voltage range only the actual air volume is increased across the range. It seems on the surface that all a tuner would need to do is change the fueling tables associated to match. What am I missing?

[robertisaar]

I will do my best to get my tune done by the end of next week with some logs so there is something to start with.

i wouldn't expect super fast progress... then again, i don't know, most of it would be copy/pasting the same code i wrote for nAst1.

I don't currently have an engine that runs on a P66 PCM but I do have the PCM and I have a chip burner that burns the 27SF512 chips and others so I don't know if that burner would suffice.

I've kinda been wanting for a while to try setting up an engine test stand to run an engine out of a vehicle. Would be a cool way to test changes without putting a DD in danger.

i want to say the PROMs are intel 28F256 pieces, but it's been a while since i've looked at them. a BURN2 won't handle them, a MCUMall Willems will though, along with high dollar units.

i've been wanting to make even a water brake load cell and really basic sensors for the longest time, really even anything that can provide a consistent load against the engine(or engine/trans/rear end?) would be great for testing even minute stuff.

[robertisaar]

It seems on the surface that all a tuner would need to do is change the fueling tables associated to match. What am I missing?

that's the hack way of doing it when you have no other alternative. SO many things will be messed up by the PCM if it thinks you're at a barometric pressure of 50kPa when you're at sea level, almost everything in the calibration would need changed to be correct.

what i do is emulate/simulate a 1BAR sensor for everything that the stock code does, then when dealing with anything above 100kPa(or barometric in some instances), boost corrections can start being done(fuel being added, spark being reduced).

i seem to remember the E-side of the P66 V6 having a BUNCH of free space available, so multiple, high resolution tables can be done.

[caffeine]

Let's get to it

[mrogersmike]

Let's get to it

Gotta start somewhere
Meant sometime, may as well be now

[robertisaar]

Let's get to it

Gotta start somewhere
Meant sometime, may as well be now

quick look: E-side without any of my previous patches has from 9381-97FF and F154-FFAD as open space. that's a combined 4,823 bytes to work with.

for the MAF side of things, the E-side TICs are(in order), 3X crank sensor, cam sensor, VATS, and then TIC4/TOC5 is configured as an input for the 24X crank sensor. so, the VATS TIC can probably be reconfigured as a pulse accumulator for a high-frequency MAF but i assume that there is a hardware low-pass filter that will need dealt with as well, so it is likely that a surface-mount resistor and capacitor will need changed to do that.

TOC1 and TOC4 also appear to be open in the code(but no idea if their circuits actually go anywhere on the boards), so that leaves two potential variable frequency/duty cycle outputs for whatever purposes one might have.

too bad their isn't any free space left on the T-side.... all of the TICs are taken(3X, cam, VSS, 24X), but there are 3 free TOCs there too...... the only TOC it uses is for its own 160Hz interrupt timer. i have to wonder how painful it would be to remove the 32KB PROM from the T-side(at least) and stick a 64KB version in and then try to keep all of the chip-selects for the other stuff on that side working correctly. a new bootloader would be needed at a very minimum, something i've never excelled at doing since it requires knowledge on the PC side of things.



so, to get the easier of the two things done first, what would we be looking for in terms of 2/3BAR support for fuel/spark?

[caffeine]

Something like what nAst1 has would be perfect, plus a port of the 2-step stuff . Any idea if this code would support a 4-cylinder as well?

[caffeine]

I guess if there's room a VE table for 100+ kPa would be ideal but the multiplier has always worked well for me.

[robertisaar]

Something like what nAst1 has would be perfect, plus a port of the 2-step stuff . Any idea if this code would support a 4-cylinder as well?

so, 3 giant tables for off-boost and the boost fuel multiplier and boost spark reduction? or what i planned on doing for newer versions that has two off-boost and two on-boost to get rid of the fuel multiplier and spark reduction tables?

the way this PCM handles dwell control is different from what i remember of when i went through all of the code. i'll need to see how different, but i'm sure it could be done too.

4 cylinder? probably, depending on what kind of signals are fed to the PCM and what kind of signals the ICM wants(the later(94+?) quad4 ICM, for instance... has two EST circuits so that the PCM is basically in direct control of the coils, that MIGHT work but would require a lot more thought than a simple 2X crank signal, single EST output would need). there are a few places in the code where the "cylinder to be fired" value is checked against the number 7 immediately after being incremented, if it shows 7, it resets to 1. could change that to showing 5 and that would take care of a lot of the quirks of running a 4 cylinder on a 6 cylinder SFI PCM. the rest are basically raw reference period related and would require changing some of the math to reflect 4 cylinders instead of 6.

I guess if there's room a VE table for 100+ kPa would be ideal but the multiplier has always worked well for me.

i figured it must work well enough, the TGP code works really similarly to what i ended up with for fueling modification. stock TGP calibrations tend to run badly but reworks tend to fix most of their issues.

i discover a small error i made for the fuel multiplication calc, the effects of it can be hidden really well in the VE/multiplier setup, but i'll eventually fix it.... it has to do with the injector dead-time offsets, the more accurately you input those, the less noticable it is, but any error there is basically multiplied while in-boost. as long as you start a calibration with a cautious amount of overfueling while in boost and dial back the AFR from there, you would never know about it unless you disassembled the patched BINs and saw what the boost multiplier table interacted with.