LS1 Boost OS V3

They go by many names, P01, P59, VPW, '0411 etc. Also covering E38 and newer here.
User avatar
vwnut8392
Posts: 59
Joined: Fri Feb 28, 2014 7:38 am
cars: AAN powered 83 audi 4000 quattro
1983 audi UR quattro
1992 GTI VR6

Re: LS1 Boost OS V3

Post by vwnut8392 »

bubba2533 wrote:
vwnut8392 wrote: over all good work on this! was wondering if you had tested your patched BIN files with HPtuners to see what happens? though i do like your solution as it is semi open source i was just curious if HPT tries to read one of your BIN files if it freaks out or doesnt care and allows it to work plus does all the checksums correctly. i know there where tuners in the past that liked to modify how the checksum worked so that if someone modified their file and tried to correct the checksum with conventional ways it would create an error and make a brick of the ECU on purpose.
Thanks. It’s not open source, but just open for suggestions. HPTuners likely would not work as-is due to the custom operating system ID. I’m guessing it would read the bin file but probably not have any tables displayed, but I could be wrong as I haven’t tested it. That’s an interesting way to protect your IP, but with this ECU it’s fairly easy to bypass almost any security method if you have the desire so I didn’t implement anything like that.
vwnut8392 wrote:anyone know what the password is so the XDF can be edited?
I won’t be giving out the password.
thats fine on the password. would have been an easier road to disassembly on it but i'll do it the hard way. as for HPtuners you are right, it freaks out and wants to upload the file to their servers after it reads it because it doesnt understand it. yes you are right about bypassing security, i would just flash an ECU with a file than lift the flash off the board and read it with an eeprom programmer if necessary. thats how i got to look at alot of HPtuners modifications to the code in the past.
User avatar
vwnut8392
Posts: 59
Joined: Fri Feb 28, 2014 7:38 am
cars: AAN powered 83 audi 4000 quattro
1983 audi UR quattro
1992 GTI VR6

Re: LS1 Boost OS V3

Post by vwnut8392 »

after looking at disassembly of your V3.4 i have a couple suggestions. why are the constants/tables after your code? this would make the user have to do full write of the ECU every time with PCMhammer. why not move them into the calibration area instead so than just the calibration alone can be written a lot faster when making adjustments/tuning the variables. with the VE table for speed density i really dont see a need to have individual tables for each MAP sensor size. could just rescale the MAP pressure axis instead for 1 table and save some room. also the original VE table is plenty sufficient in size so its pressure axis could be rescaled to suit each MAP sensor. Im sure there are tables and constants that can be disabled or re-purposed in the calibration area to make room for all of the additions. like instead of having a high octane and low octane ignition maps why not make the ECU just read the high only and reuse the low octane ignition map for a VE table if you want larger axis's for more resolution?

also it it is very possible to auto calculate the MAP pressure axis automatically based on bit pair in the BIN file. like pick input 0 that would work for 2BAR MAP sensor. 1=3BAR MAP sensor and 2 for 4BAR MAP sensor and after that your axis displays properly for that sensor. i dont have time to elaborate on the specifics on how to completely do it but it can be done. with a method like this implemented you only need 1 VE table that does everything and eating up less code space.

another idea would be to make the user pick what MAP sensor is being used in your patch program and it inputs the right axis for that sensor on all of your tables.

also the choice of units of measure in the XDF would be nice as well. to make most if not all of your pressure based axis's and pressure based constants to read in PSI relative to atmospheric pressure like us americans like to use you can just modify the current equation like this.

Code: Select all

(YOUR CURRENT EQUATION)*0.14503773800722-14.6959494
just put the current equation in brackets than multiply by *0.14503773800722-14.6959494 and it will display perfectly in PSI relative to atmospheric pressure and stay 1 to 1 with the old equation. for example if your at 0 in PSI relative it would be exactly 101.32500411216033 KPA.

your on to something here but it just needs to be refined more.

lastly i changed the OS ID back to stock in one of your patched BIN files and HPtuners has no problem with it after that. works totally fine.
kur4o
Posts: 948
Joined: Sun Apr 10, 2016 9:20 pm

Re: LS1 Boost OS V3

Post by kur4o »

vwnut8392 wrote:after looking at disassembly of your V3.4 i have a couple suggestions. why are the constants/tables after your code? this would make the user have to do full write of the ECU every time with PCMhammer. why not move them into the calibration area instead so than just the calibration alone can be written a lot faster when making adjustments/tuning the variables. with the VE table for speed density i really dont see a need to have individual tables for each MAP sensor size. could just rescale the MAP pressure axis instead for 1 table and save some room. also the original VE table is plenty sufficient in size so its pressure axis could be rescaled to suit each MAP sensor. Im sure there are tables and constants that can be disabled or re-purposed in the calibration area to make room for all of the additions. like instead of having a high octane and low octane ignition maps why not make the ECU just read the high only and reuse the low octane ignition map for a VE table if you want larger axis's for more resolution?

also it it is very possible to auto calculate the MAP pressure axis automatically based on bit pair in the BIN file. like pick input 0 that would work for 2BAR MAP sensor. 1=3BAR MAP sensor and 2 for 4BAR MAP sensor and after that your axis displays properly for that sensor. i dont have time to elaborate on the specifics on how to completely do it but it can be done. with a method like this implemented you only need 1 VE table that does everything and eating up less code space.

another idea would be to make the user pick what MAP sensor is being used in your patch program and it inputs the right axis for that sensor on all of your tables.

also the choice of units of measure in the XDF would be nice as well. to make most if not all of your pressure based axis's and pressure based constants to read in PSI relative to atmospheric pressure like us americans like to use you can just modify the current equation like this.

Code: Select all

(YOUR CURRENT EQUATION)*0.14503773800722-14.6959494
just put the current equation in brackets than multiply by *0.14503773800722-14.6959494 and it will display perfectly in PSI relative to atmospheric pressure and stay 1 to 1 with the old equation. for example if your at 0 in PSI relative it would be exactly 101.32500411216033 KPA.

your on to something here but it just needs to be refined more.

lastly i changed the OS ID back to stock in one of your patched BIN files and HPtuners has no problem with it after that. works totally fine.

I will try to get you some answers without being affiliated with the boost OS in any way.

You need full flash only once to update the checksum addresses. After that it is partial flash with only the segments that needs changing, usually the stock cal segment and the new patch segment.
With almost half the chip blank you are not really concerned about space, so no need to shrink anything to fit, you can even fit almost a second OS there. Program will write only the new segment, and trying to fit that much data in existing calibration area is not worth it. That`s why the 512kb pcm just drop out of the box, not enough space for all the goodies.

Hp tuners can work but if the checksum areas are changed you are looking for troubles.

For scaling and code optimization only bubba can give you more info.
bubba2533
Posts: 498
Joined: Wed Apr 11, 2018 8:50 am
cars: 03 Chevy S10 Turbo V6

Re: LS1 Boost OS V3

Post by bubba2533 »

I think this is a better discussion for my development thread so I'll reply over there.
LS1 Boost OS V3 Here. For feature suggestions post in here Development Thread. Support future development ->Patreon.
bubba2533
Posts: 498
Joined: Wed Apr 11, 2018 8:50 am
cars: 03 Chevy S10 Turbo V6

Re: LS1 Boost OS V3

Post by bubba2533 »

Keeping for posterity:
So it looks as thought there is a bug in the closed loop fueling code. It works reasonably well at idle and cruise, but there is something I will need to do more testing on to figure out what is going on with wide-open throttle.

I would recommend not using it until there is a fix.
Edit: It looks like this is limited to V4 as the closed loop code was optimized in the process of doing other things for V4. Should be a simple fix for V4 so an updated version should be released soon, but I want to do some testing first.
LS1 Boost OS V3 Here. For feature suggestions post in here Development Thread. Support future development ->Patreon.
gilius
Posts: 4
Joined: Wed May 31, 2023 6:07 am

Re: LS1 Boost OS V3

Post by gilius »

Bubba, thank you for the hard work with Boost OS. I have a couple of questions. On Wideband Closed Loop Fueling. is your patreon version going to be able to monitor two oxygen sensors on both banks? Will the flex fuel tables have a higher resolution than what they come with stock? The stock flex fuel tables seem to be unusually coarse compared to standard fuel tables on the P01. Also, have you posted any further updates on your P59 that you modified with the daughterboard?
bubba2533
Posts: 498
Joined: Wed Apr 11, 2018 8:50 am
cars: 03 Chevy S10 Turbo V6

Re: LS1 Boost OS V3

Post by bubba2533 »

gilius wrote:Bubba, thank you for the hard work with Boost OS. I have a couple of questions. On Wideband Closed Loop Fueling. is your patreon version going to be able to monitor two oxygen sensors on both banks?

Thanks! That's the first time I've gotten that request. It's possible to change the code to add that, but it would use another 0-5 volt input. There aren't that many 0-5 volt inputs available, and I would have to decide which one to use. If there is a decent amount of interest I could add it. I just started a discussion with all my Patrons about what I should include in V5, so I'll add it to the list of possible options.
gilius wrote:Will the flex fuel tables have a higher resolution than what they come with stock? The stock flex fuel tables seem to be unusually coarse compared to standard fuel tables on the P01.
I'm not sure what you mean by flex fuel tables. I've had no issues with the flex fuel in my truck.
gilius wrote:Also, have you posted any further updates on your P59 that you modified with the daughterboard?
Not yet, but I'll get back to it at some point soon hopefully.
LS1 Boost OS V3 Here. For feature suggestions post in here Development Thread. Support future development ->Patreon.
gilius
Posts: 4
Joined: Wed May 31, 2023 6:07 am

Re: LS1 Boost OS V3

Post by gilius »

bubba2533 wrote:
gilius wrote:Bubba, thank you for the hard work with Boost OS. I have a couple of questions. On Wideband Closed Loop Fueling. is your patreon version going to be able to monitor two oxygen sensors on both banks?

Thanks! That's the first time I've gotten that request. It's possible to change the code to add that, but it would use another 0-5 volt input. There aren't that many 0-5 volt inputs available, and I would have to decide which one to use. If there is a decent amount of interest I could add it. I just started a discussion with all my Patrons about what I should include in V5, so I'll add it to the list of possible options.
I don't mind making a sufficient contribution to Patreon to encourage this happening. It's sorely needed. Imo it's indespensible to be monitoring both banks on wideband, especially considering that alot of racers are monitoring each cylinder independently with highly custom equipment. In that regard, two bank monitoring is a relatively basic requirement that is possible with available hardware. I find it fascinating that the stock computer can be made to do this; it seems reminiscent of the numerous advantages that open source code and approach has brought to the masses.
gilius wrote:Will the flex fuel tables have a higher resolution than what they come with stock? The stock flex fuel tables seem to be unusually coarse compared to standard fuel tables on the P01.
bubba2533 wrote:I'm not sure what you mean by flex fuel tables. I've had no issues with the flex fuel in my truck.
I should be more specific. In the P01 you have this kind of fuel table:

Image

In the P59 you have this kind of fuel table:

Image

The P01 is a non flex fuel table. The P59 is what I think of as a flex fuel table. Resolution is the problem; P59 has 15 to 100 kPa, but the P01 has 5k increments all the way to 220k, then the increments start going to 10's. Sloppy Mechanics calls this a fantastic amount of resolution. Comparing the P59 table to the P01, Sloppy notes that this is essentially 1/3rd of the resolution.

With higher capacity injectors, (1650cc in my case) and speed density tuning data being crushed down to 1/3rd of the resolution that it could be at, you run into problems. Sloppy has run tests on this and the best thing to do here is probably let him describe his own analysis:

https://youtu.be/pPDkrnxaMVc
gilius
Posts: 4
Joined: Wed May 31, 2023 6:07 am

Re: LS1 Boost OS V3

Post by gilius »

Looks like I'm late to the party. I joined a few weeks back and yesterday started going over the Boost OS dev thread. If I am understanding things correctly, the fuel (aka VE) table problem was already fixed a while back?
gilius
Posts: 4
Joined: Wed May 31, 2023 6:07 am

Re: LS1 Boost OS V3

Post by gilius »

Bubba -- Patreon joined
Post Reply