Seriously this thread.

It may be alive lol

Final planning/parts ordering for 98 LS1 running 99+ 411 in the 91 Formie.

My head is going to hurt sorting this out now

 

will this kind of USB cable work ?

ELM 327 interface v1.4a
Protols that can be used:
- J1850 PWM (41.6 kbaud)
- J1850 VPW (10.4 kbaud)
- ISO 9141-2 (5 baud init, 10.4 kbaud)
- ISO KWP (5 baud init, 10.4 kbaud)
- ISO KWP (fast init, 10.4 kbaud)
- ISO CAN (11 bit ID, 500 kbaud)
- ISO CAN (29 bit ID, 500 kbaud)
- ISO CAN (11 bit ID, 250 kbaud)
- ISO CAN (29 bit ID, 250 kbaud)

i'm European and looking for correct cable on my domestic market

best regards

 

Apparently the ELM based cables will not work for flashing the PCM. The cable needed is mentioned earlier in this thread. Its quite a bit more expensive than the ELM's but this is still a MUCH less expensive solution than bending over for HPtuners or their kind.

 

ok, thanks for info and i'll have to order one

best regards

 

Got the same type of project about to start.....

1996 S10 extended cab + 40K mile 2012 5.3/6L80E + 2012 Silverado PCM & TCM

Will be running fly by wire throttle and keeping as much of the factory 5.3 wiring as possible.

Will be adding in a cam and shorty headers to the 5.3. Most likely will end up with a factory takeout 6.2 Camaro cam. Will be keeping the variable valve timing and displacement on demand. The truck will still have a cat on it and likely end up with a 3" single exhaust for stealth.

The truck currently has a 3.08 open differential, but I have a 3.42 posi hanging around from a F-car that I plan to swap in.

The truck many end up with a body lift to help clear the transmission as well as a mild suspension lift for some larger A/T tires. I know it will be receiving a corvette radiator in the forward mounted location and my brother wants to keep his A/C and towing capacity. It will likely end up with a mechanical fan and the engine set slightly rearward toward the firewall. How far depends on the 6L80E clearance to the tunnel, which may receive some gentle persuasion with a big hammer.

Our goal is to keep as many factory parts as possible, building a nice daily driver/work truck, that gets 2.2L like fuel economy on the highway and 2x the power of the original 4.3.

 

The rotor tip is pretty close to the base circle of the terminals, but since they have all those traces inside the cap for the side terminals, a large spacing is needed to route everything inside. Think its something like a 5.25" circle. Obviously the larger the gap the spark has to jump, the higher the voltage the system is exposed to. The short high voltage pulse ahead of the plug firing forms corona breakdown in the dielectric, and then the main pulse from the coil forms a carbon track when it travels along the corona path.

 

ELM Cables dont support large continuous block transfers, which some of the PCMs require to properly enter reflash mode (basically you offload reflash firmware to the PCM thats several kB in size as one frame). The ELM also does not have teh 4X VPW protocall. The AVT cables support pretty much everything.

 

ok, can you give specification what requiments must be met by cable to be fully functional? then maybe i'll be able to find even cheaper solution

best regards

 

This mean anything to anyone?
http://pastebin.com/LXcS3C83

Supposedly some seed key info here demented...

Looking into nabbing the whole 256 seed key table

 

J1850 VPW 1x and 4x.

NExiq USB would work if you can get the commands and such for it and dimented wants to add it as a supported device.

A cheaper solution that does VPW can and 1939 and only that would be ideal to keep costs down. the AVT devices are super overly robust.

 

Any further updates on this any of this? 411 support, etc? Following this has been pretty exciting so far!

Thank you for the hard work guys!

 

I too am following this thread closely. Just picked up an LQ4 to build for my obs suburban. The Vortec with a burnt valve is coming out now since last week the transfer case went out, which has accelerated my swap plans. So I'll just do the swap and fix the trans all at once.
I guess I may order the adapter to program the 28f400 directly in my burner sometime in early June if this is not ready yet by then. But I'm still hopeful about it. and even if it takes longer I'll just plan to get an AVT later. Between myself and some friends/family members I have about 7 or 8 projects that would be better with an 0411 swap.

I'm very confident that Dimented will come through on this. Also Very gratefull, this forum and the people on it are the best.

 

Still working on getting requisite can info. Doing some work figuring out how another tool does it. I have what I think to be a complete seed/key table, now I need to figure out what commands it sends to the avt. Not in the code anywhere. it must be in a dll I guess

 

I finally finished my camaro and figured out how to get rid of the dreaded P1336 DTC that haunts just about every new PCM swap, but thats a story for another thread. The LS1 PCM is next in the project list. Hoping to start playing with it this week

 

 

I must have just gotten luck when I swapped the L31 engine in my Express and swapped to the 0411 reflashed from an 02 5.3 Tahoe. Never have seen a P1336 code.

 

My first PCM with a ZIF socket never had a P1336, but when I swapped over to an unmolested one, I got the P1336 immediately. If the PCM sees enough variance between what it has stored and the actual crank pulses or it looses its eeprom, it will set a P1336. This happened to a buddy at work when he tried to jump his Impala. The PCM lost the stored learned parameters and set the DTC. It also ran like crap, so it can effect performance if its off enough.

I didn't notice a huge difference after doing the CASE relearn, tho. Still ran good even with the P1336 set, but it did seem smoother at WOT with the proper CKP variance.

 

I have since set the CMP and performed a CASE relearn but I saw no noticeable change from either. My CMP was set within 2* without a scan tool, just my eyeball.

I have seen other 0411 swaps that were barely driveable until the case relearn was performed.

 

Dimented24x7,
I have been thinking about upgrading my $0D code and installing a 411 PCM. I want COP ignition.

1) is there a 411 code (OS) hack out there or a start of one? Is so, got a link?
2) I am going to look into using TTDigger (freeware) as the disassembler. Is the OS code finally a standard thing with just a CAL difference per vehicle?
3) Not sure on the assembler yet
4) Writing some C code for an AVR chip to do the VPW 4x rate (emulate a DLC).
5) Look at the J1850 bus signal conditioning. For prototyping, maybe pull apart a ELM327 device and install AVR in place that is 4x capable.
6) I read block mode is required for flash updates. Therefore, 4x mode is required. Do you know the block length required? That will determine amount of calibration cycle needed, if any.

My first step is doing the DLC emulator. If I can't get past this then I will probably go the desolder flash, install socket route.

 

1) is there a 411 code (OS) hack out there or a start of one? Is so, got a link?
I posted the hack to the '411 on this message board. A search under my posts should bring it up.
2) I am going to look into using TTDigger (freeware) as the disassembler. Is the OS code finally a standard thing with just a CAL difference per vehicle?
Only for the 2002+ PCMs. Earlier ones have several OS flavors.
6) I read block mode is required for flash updates. Therefore, 4x mode is required. Do you know the block length required? That will determine amount of calibration cycle needed, if any.
The device may need to send up to, and beyond, 4kB of data. The reflash requires that you supply your own interface software thru the OBD as no reflash firmware is supplied within the PCM. The PCM also expects that the user executable be sent all as one block, so the device will need to send at least 2-3k, but the reflash software I wrote was in excess of 4k with all the features I needed. I think the LS1 PCMs do not specifically require that 4X be used to send the software, but its a lot quicker if you can. In 4X mode sending several kB of data takes only seconds.

 

I have VPW 1x working on an atmega328p (2K of ram) that can be compiled for a atmega2560 (8K of ram) chip. It communicates with the PC at 115.2K RS-232.
It works, but much more work needs to put put into the noise filtering and waveform shaping for 1x mode. Not a problem in 4x mode.

 

If this can be done with an AVR that would be a dream come true for me. I have a drawer full of AVRmega's and some XMega's too. As well as a PCB printer (homemade from a modified inkjet).

 

Working the 4x mode. I had to do a re-write of some of the 1x mode when I changed the HW scheme. Also, working simulations for the waveshaping.

Did some thinking about 4x block mode buffer size. RS-232 115.2K is streaming at 8.6 Usec per bit, and VPW 4x is at an average of 24 Usec per bit. In theory, no buffers are really necessary under one condition. That condition is, the PCM flash utility must have a CRC on the 4K-ish subroutine that it loads to ensure that there were no transmission errors on both the UART interface and the VPW interface.

If the subroutine to the PCM was not streamed, it could still have TX errors on the VPW side and requires a CRC.

Overall, I don't see why having a large 4K-ish buffer on the VPW device is necessary if the PC can send the receive/transmit at an "extremely slow PC rate of 115.2 Kbps".

In saying that, using the atmega328P would allow for a 1k-ish RX and 1k-ish TX buffer.

What it really comes down to is how existing PC software is written to use the 4x block mode. If it does not stream and simply does a "somewhat of a burst" with small time gaps in between bytes then I have to support that and buffer the entire 4128K bytes of block data before transmitting over the VPW bus.

EDIT:
Decided to create a "packet mode" that should work with existing software, and have a "stream mode" that allows essentially unlimited sub-routine size depending on the PC USB rate. If the PCB USB rate is 115.2K then you could stream an unlimited size data, but will limit the byte count size to something like using 4 bytes (32 bit) so a packet of 2^32 = 4Gig could be transferred in one command.

Complete re-flash program time at 4x block mode:
24 usec_per_bit * 8 * 512K ~= 2^5 * 2^3 * 2^19 usec = 2^27 usec = 128M * Usec = 128 sec -> 128/60 ~= 2 minutes

Complete Calibration program time at 4x block mode:
24 usec_per_bit * 8 * 96K ~= 2^5 * 2^3 * 2^17 usec = 2^25 usec = 32M * Usec = 32 sec

 

So does this eliminate the need for the AVR cable. I would imagine this is your own hardware design to flash the pcm? If so is this something you are going to be releasing a DIY parts guide and what not. How much will something like this run to make? Pretty interesting the things you guys can do. Keep up the progress. I would love to be able to convert 0411s to ls or express van from a v6 application at will. Then they can be tuned with an unlimited license with EFIlive/hptuners or when proper definitions are finished with TunerPro.

 

Neruve,

Junkcltr is talking about programming an Atmel AVR microcontroller as the brains of a new interface cable, as an alternative to the very expensive AVT cable. This thread is mainly about dimented24x7's re-flashing software that is designed for the AVT based programming cable (fancy commercial cable).

The AVR microcontroller is a self contained single chip computer that cost around $1-15 depending on which version you need. They operate at 8,16, 20 or even 33Mhz and execute around 1 instruction per clock (very fast for what they are). Most AVR models have a DIP based version that can be easily be soldered onto a basic "sea of holes" type project board (radio shack for a few bucks). If you start researching microcontrollers you will discover the "Arduino" which has a HUGE following right now, this is really just an AVR chip programmed with a simplified version of the "C" programming language.

Generally for a really basic AVR type microcontroller board you need at least a 5v 7805 voltage regulator, a few resistors, a crystal, a couple of small capacitors (so the crystal can oscillate), a small breadboard, wire, solder, small enclosure and of course the AVR chip itself. Probably under $20.00 if you have nothing other than a soldering iron to start with. If you do these very often then your costs drop dramatically since you'll usually buy small components in packs of 10,50,100 for about the same price they sell a 5 pack at the shack.
If building an AVR based interface your most expensive single component will be an OBDII plug. But of course this would be DIY which isnt for everybody. The existing AVT based cable will set you back at least a couple hundred, but is said to be "very robust".

Neither option is ready to go right now, I'm hoping both paths to re-flashing become a reality soon. I'm sure HPT and their kind really dont want to see this happen, ever.

Marvin

 

Marvin,

Thank you for the info. I have looked into so Audrino stuff before but never really did anything with it, although I would like to.

In the parts list supplied the only thing I am not familiar with is the "crystal" not really finding much while searching either. Are these called something else or is there model or number I could look at for a visual example? Would you be able to use like a 16mhz resonator instead? This may eliminate the need for some of the caps.

EDIT: Never mind, obviously i'm an idiot as I found the crystal stuff with a quick search, must have been because the first time I searched was early morning and I was half asleep...

All of this is pretty interesting. In theory, once something like this is released, a DIYer, along with correct TunerPro defs, (seems like developing defs for the 02 12212156 OS would be ideal, as you would be able to flash a stock bin from the particular application you are working on, be it ls, express van gen I, or v6 and tune accordingly) could pretty much put a 0411 on anything and not have to pay an insane amount of money to do so.

I do believe this may be a blow to the main stream solutions, I can't imagine them losing a whole bunch of money, figuring this is only able to do a few ecms, and they are branching into the newer stuff, shops and what not, will still use the supported solution, therefore the people that would be willing to try to make something like this or, if need be, spend 300 bucks on the other cable will just be people that like to do things on their own.

 

Still some work to do for the 02 cals as they are based on the universal OS, which there are no XDFs for yet.

Also, still working on the app for the '411 flash. The OBD-II DLC (data link controller) in these is much different, so thats a little bit of a sticking point to get thru. I think some of that may have been why the previous effort to make a flash app failed. There are some things like 4x mode and DLC interrupts that are controlled via hardware I/O instead of DLC commands. If the DLC is still processing interrupts, it does so transparently, but still can interrupt whatever process your working on, like flashing. The previous PCMs had a simple off the shelf M68HC11 DLC which were easier to use. These are custom and not as easy to configure...

 

So I was at the local pick-n-pull a couple weeks ago and picked up an 0411 from an s-10. I saw several similar computers (same case) but they had a different colored plugs. Some color variations I seem to remember are green/blue and clear/??? .
I've been searching around to find some kind of compatibility list or chart to show which PCM's can run which OS and what their capabilities are and required hardware etc... so far I'm not finding very much useful info that is aimed at retrofitters such as myself.
I did find this little quote someplace;

Quote:

The service codes for '03-'07 include 12576106, 12586243, 12586242, 12583560, 12583561, and 12589463. These PCMs are interchangeable, with a complete re-flash of the appropriate operating system.

I suspect that I passed on quite a few PCM's that were possibly usable on one of my projects, but I had no way of knowing. Does anyone have such a chart of PCM's already?

 

Yes I have wondered myself if you could, for example, take a 99 silverado PCM and flash it with the 01 or 02 file from a 0411 or if there is an issue with this as otherwise it looks identical. I know pinouts for bench harnesses in the efi live guide is 1999-Current LS1, LB7, Early V6 and 4 Cylinder Style Controllers. I know that doesn't stretch anywhere close to current, however, I'm wondering if any other 4 or 6 cylinder pcms around 99-02i-sh can be flashed to something of a v8.

Also wondering if junkcltr has made any progress with his project in the past week or so.

 

Actually I just found this last night, while searching for something else of course.

I think it is a start at what I was thinking of. I wish I could remember where it came from.

 

EFI Live forum has a thread with updated files, not sure which is latest version...

 

Thanks for mentioning it was on EFI live, this currently seems to be the latest and greatest version.
post #95
https://forum.efilive.com/attachment...0&d=1363970440

Just looking at it I learn something new, the L21 454 engine starting in the late 90's had ls1 style cnp like the 8.1. a quick search confirms this. Possibly another parts source for one of my big block projects.

 

Making progress. Using a motorola transceiver for wave shaping instead of home brew. I am focusing on the code portion instead of analog hardware right now.

I have the 4x block transfer code written and testing now.
Supports up to 4128 bytes at 4x speed.

 

Awesome, I check this thread every day to see where this is going.
For many, this will be a game changer.

Marvin

 

I also check this thread everyday. Very exciting to see people making progress with this.

Between dimented, antus at pcmhacing, and junkcltr someone has to be real close to a solution. I would think it should be coming up soon when someone has a solution. $300 AVT cable or not.

 

Sorry to keep everyone waiting, lol! Been a while since I had any time off...

I found out almost all of what I need to know. Going to try updating the black box code with the stuff for the '411 this weekend. Theoretically, it should work once the DLC logic and flash Vpp are updated, but its still a decent chunk of work to bring it up to speed.

 

dimented, Is the cable that comes with the Jet dynamic spectrum tuner the same AVT cable that you are developing this software for? As in, if one has that will it work?

 

I think it may very well be. AFAIK, Tuner Cats uses the AVT, or something that looks like the AVT cable. You can tell for sure by hooking the cable up and opening it in a serial Terminal program. Send the hex value 0xF0 thru the terminal, and if its the AVT, it will respond with its model number in hex format, something like 0x08 0x52, or 0x08 0x42 if its an AVT cable. I attached a simple freeware terminal program that works well for quick communication check over the vCom port.

 

Also, if anyone has any of the seed/key algos for the various 01-02 PCMs, I would appreciate contributions

I can probably back work one or two of them, but it will only unlock a limited number of PCMs that way.

 

It is the same cable. But somehow programmed differently as you need licenses per PCM... I've been told it is in cable and not sftware but never checked myself.

 

A quick check is to manually send it commands using a terminal program. AVT site has the users manual that has the AVT command set in it. If you can do basics thru the terminal, then my program and tunerpro should also work with it.

It does have an ARM in it, so its not out of the realm of reason that it has custom firmware for the applicaiton.

 

I don't have the info on the algos, but I was wondering about doing the following.
What if the VPW cable had a command you could send it to toggle an output.
That output would be connected to the power supply pins into the PCM.

The sequence would be:
1) PC sends the power-up command to the PCM
2) PC then gets seed and guess at the key
3) PC sends power down-command and waits for caps to discharge, then goto "1)"

You could do this loop at worst case 500ms. So 64K *.5sec = 32K seconds which is about 9 hrs. Best case is probably 250ms which equates to 4.5 hrs. Those are worst case needing to do 64K guess.


Still working the VPW 4x cable. Got sidetracked with some $8D boost code.

Dimented24x7,
Does your code require that the cable ID response must be an AVT cable ID?
Does you code use the periodic message function of the cable?
Would you have a list of the cable functions (commands) that your code uses?

 

I thought somebody cracked the "seed key" thing a while back?
oh well, sure is good to see somebody able to work on this.
The $8D boost sounds interesting also.

 

See my responses below in blue.

 

Yes, but I haven't seen all of them. They might be posted, but its been some time since I really dug for them. Will have to confront this at some point if the LS1 reflash tool is to be useful across multiple platforms...

 

I was thinking that maybe this might be more useful to people if its done as an open source collaboration. Ive been getting questions to add support for this cable or that PCM, all of which could turn into a full time job by itself!

If someone has a site they are willing to give some space on, I can post up the source code as well as info on the PCMs and hacs so people can do whatever they want with the interface.

At least this way I wont be the bottleneck, lol. At the rate things are being piled on at work, I'll never get around to actually doing any fun stuff any time soon

 

I was just thinking the same thing in terms of source release and maintenence last week. One option is to put it on sourceforge.net. It can be checked out from there.

In terms of the cable, the current path is to implement the functions required to do the tasks. The only problem I saw was having a cable that needed to report it is made by another manufacturer, but you answered that question.

I can help with this once the cable is complete. I have the 4th gen already so it isn't an option of being able to tune it, it is a matter of when. I am getting back to the VPW 4x hardware in a week.

 

Any updates for your anxious followers?

 

Bump. Any good news?

 

Great work dimented24x7! Iv reading up on your threads, especially "pcm hacking 101", a bit of a steep learning curve but nevertheless a great tutorial!

Let me know if you need the LS1 algo still. I have that coded up and ready to go.