[rjahl]

I run most of this software in Virtual Machine running a Windows XP shell on top of my Windows 7 64 bit. Nice thing about the Virtual machine is you can keep copies of varies set-ups. I created a blank XP with Basic software and saved that as a base. All other installations are saved as separate VMs that I can mess with or delete as I please. Each VM takes up about 20 gigs of HDD space, nothing by today's standards. VM ware has a free version that you can download directly from them, Personal use only.

The big pain I have with Inpa, Winfkp, Tool32 etc is the various versions available. Some are old, in English but lack coverage for my car. New versions that cover my car all seem to be in German. Just can't seem to get both. I gave up and settled on the German versions.

[hassmaschine]

yeah. some of the newer ones are missing features older ones have as well. Mine are all in german but you get used to it. a lot of the translations are missing context anyway.

The only one I need to update is my NCS Expert because I can't code my E90s.

VM's are nice - I mainly have the physical machine because I don't like hauling my nice laptops around in the car. Plus it has real serial ports which just work better for some things. Just don't go on the internet with it.. lol

[rjahl]

Ha ha ha, I did it, I have inserted the 330/3.0si code into my spare dme and its running without codes, running very strong!! Too strong to really check tonight , it's raining and the extra pull is a little scary. I hope this does not have me looking into LSDs.

I guess this proves the point that you can run the higher power class software in a 325,328. Forget the powerclass setting and the link to your VIN. That is breakable.

I am running 99.99999999999 percent of the original code. I guess the next step is to prove how it can be done with winfkp. I'll seek advice from the forum regulators before I share how I did it. I'm not looking to create trouble.

[hassmaschine]

Just share it. There's nothing the tuners know that can't be figured out comparing stock files. How do you think they came up with the 3 stage tune...

Actually started working on it a bit today. Played with some of the disa transitions, pretty happy with the tune now.

At one point though it seemed down on power (it felt stock). Well, i was changing the oil and popped the engine cover. The #2 coil had backed out. It was still so strong i wasn't sure there was really a change. Plugged the coil back in and wow.

Yes, you will want an lsd. I have to use dtc all the time or it drives me crazy, but the peg leg still sucks. BMW marketing can say the electric aides make it work like an lsd all they want, but it just isn't the same. Cost cutting...

[Taskmaster]

Originally Posted by rjahl

Ha ha ha, I did it, I have inserted the 330/3.0si code into my spare dme and its running without codes, running very strong!! Too strong to really check tonight , it's raining and the extra pull is a little scary. I hope this does not have me looking into LSDs.

I guess this proves the point that you can run the higher power class software in a 325,328. Forget the powerclass setting and the link to your VIN. That is breakable.

I am running 99.99999999999 percent of the original code. I guess the next step is to prove how it can be done with winfkp. I'll seek advice from the forum regulators before I share how I did it. I'm not looking to create trouble.

Where in Tampa are you? Wanna meet up and compare the cars?

[Taskmaster]

Guys, are you telling us (me) that we can use the MSV70 DME info on the MSV80? IS there anything in the code that would explain why MSV70 cars still have 10whp+/- on MSV80 cars?

[hassmaschine]

Yes, there's definitely things that can be copied directly. The difficult thing is msv70 uses a different byte order than msv80, so comparing files is harder.

The parts that are most difficult to find are configuration bytes, because they may be in a different order or not exist at all. Curves and maps are easier, i already found most of them - but since i have an msv70 car, that has been my focus.

in the next few weeks i want to finish my msv70 xdf. Then i can start searching through msv80 again.

What we are really missing is a way to flash without an expensive, proprietary cable that locks to your car. Using WinKFP may be a solution. We also need to figure out how to calculate the checksums, right now my bdm hardware does it for me but that doesn't help you guys..

[rjahl]

Let's be clear, I drive a z4 with the n52 motor and MSV 70 DME. I'm using this forum because Hass and his progress is much more interesting than the z4 forums that seem to concentrate on the M versions. I took the parameter file for the 3.0 si (0da file) and converted it to binary. From there I inserted most of that bin file into my current flash. I exclude the beginning sections that indentify the file and I modified a parameter that seems to be identified as a powerclass byte. I left the value of that byte as 0 instead of 2. If the DME checks this byte against my VIN or anything else it will be looking for a 0 and not throw a code.

Both my tuning software and my BDM tools correct the checksum and the even agreed with each other.

I started this project hoping to use winfkp but took a short cut last night when I had already purchased the tools and had the spare DME sitting on my workbench. Now that I have a working file it should be easy to build a working ODA file. The checksums are allready sorted for this file but I know where the ckecksums are located and what data range they apply to but I'm not good enough to calculate them.

I have proved that you can run a different powerclass file in msv70 and the full conversion from a 325/328 to 330 is possible using BMW tunes. I did the 3 stage manifold swap a few month back and z4 exhaust is the same for both USA versions of the E85.

Today is full of yard work but I'll try to give more specifics in the next day or two.

[rjahl]

Originally Posted by TheAxiom

Where in Tampa are you? Wanna meet up and compare the cars?

I'm up in Wesley Chappel. Going to be a busy few weeks at the office so I don't expext much free time for a while.

[hassmaschine]

rjahl, if you give me the range I can try to calculate them. finding the location was easy, just upload an uncorrected file, then download it after the checksum is corrected and compare - but I wasn't able to figure out what the range was.

[rjahl]

Hass,

Look at the attachment,

The way I read this is the first four bytes are the checksum.

Followed by the number of blocks , 2

Ranges,
840240 - 84FFFF Block One
850000 - 85EAFF -Block Two

If you are looking at the complete Flash file, Drop the 8 on the address. If looking at a BIN from the ODA file the 8 is needed.

Each of the other checksum in the program section has a similar format but the 8 is not shown.

[rjahl]

I'm currently running the upgraded clone in the car with the original DMW setting on a shelf. How long can I expect the original DME to say synced with my car? If I try it in 6 months will I need to break out INPA and re sync it or will it always work?

BTW, the car is still running great. No rough running, knocking or any bad habits. Just how you would expect a new car to work. Except I now have a lot more power.

[hassmaschine]

I think it will always work. you would probably want to reset adaptations though.

[Mike.]

Wait so hass, in your humble opinion is creating a tranny flash possible? If you came up with one.. There is quite a large market!

[hassmaschine]

Technically yes, it could be done. The difficult part is there's not likely to be a definition file or anything for what the code in the transmission flash means, and few people doing it (or nobody) means searching for information online is fruitless. if somebody has a dump from a trans I'm willing to look at it, but depending on how it's organized it might be really difficult to decipher.

the other guys got the flash to work because they just copied somebody else's software (Alpina). They don't actually know how it works. Since we don't have the equivalent of the Alpina flash to copy, we'd be developing it on our own - and I don't have an automatic to test on.

I wonder what kind of CPU the transmission module uses. MSV70 is easy to flash because it has a BDM port, but on a transmission module you might be stuck using WinKFP. it's not like Galletto has support for random transmission modules.

[Mike.]

Originally Posted by hassmaschine

Technically yes, it could be done. The difficult part is there's not likely to be a definition file or anything for what the code in the transmission flash means, and few people doing it (or nobody) means searching for information online is fruitless. if somebody has a dump from a trans I'm willing to look at it, but depending on how it's organized it might be really difficult to decipher.

the other guys got the flash to work because they just copied somebody else's software (Alpina). They don't actually know how it works. Since we don't have the equivalent of the Alpina flash to copy, we'd be developing it on our own - and I don't have an automatic to test on.

I wonder what kind of CPU the transmission module uses. MSV70 is easy to flash because it has a BDM port, but on a transmission module you might be stuck using WinKFP. it's not like Galletto has support for random transmission modules.

I wonder if a dump of a CTS would work

[hassmaschine]

doubtful. the transmission might be the same, but the program code will be different - unless GM uses the same CAN protocol & messages as BMW does. I know they changed the format of the CAN messages on the E90 (faster, mode data). it probably doesn't even use the same computer to control it.

but, if it was the same code and the same computer, then possibly.

[Mike.]

Originally Posted by hassmaschine

doubtful. the transmission might be the same, but the program code will be different - unless GM uses the same CAN protocol & messages as BMW does. I know they changed the format of the CAN messages on the E90 (faster, mode data). it probably doesn't even use the same computer to control it.

but, if it was the same code and the same computer, then possibly.

this seems like more trouble than its worth

[hassmaschine]

do a manual swap, that'll make it like 1/2 to a full second faster without any software.

[Dan Bilzerian]

Hmm, So I think I understood 1/3 of this maybe. Are you trying to create your own tune? Or have you successfully done that.

[hassmaschine]

I've already done that. I tweaked our 325xit as well but without the 3 stage it doesn't make a big difference, plus it's slow anyway since it's so heavy.

I need to go back to the dyno with the ignition & DISA changes. I think I can eliminate the huge torque dip after 6900rpm, but I haven't been able to get any datalogging to tell me if the throttle is closing or perhaps if DISA is shutting down at the default redline. I'll go back once I'm sure I've nailed that part down.

With ignition tweaks there was definitely an increase in mid range torque but I couldn't really tell any difference in power. My ass dyno isn't very well calibrated though.

[hassmaschine]

Originally Posted by rjahl

Hass,

Look at the attachment,

The way I read this is the first four bytes are the checksum.

Followed by the number of blocks , 2

Ranges,
840240 - 84FFFF Block One
850000 - 85EAFF -Block Two

If you are looking at the complete Flash file, Drop the 8 on the address. If looking at a BIN from the ODA file the 8 is needed.

Each of the other checksum in the program section has a similar format but the 8 is not shown.

Looks right. I can confirm that 400A4 is the checksum.

I know it's a CRC32, but it might need some sort of key or seed to calculate correctly. Even selecting that range, I can't get it to match the original checksum.

What does the hex look like at 40240 on yours? Here's the string from mine:

00 00 00 02 00 84 00 00 00 84 00 FF 00 84 02 40 
00 85 EA FF 00 00 01 00 00 01 E8 C0 FF FF FF FF

00840000 is the "data area" start address in the DME memory (this is actually helpful for me for disassembly). 008400FF is some sort of DME data/code or something - don't know the meaning of any of it. 00840240 is the start of block 1 (again?) and 0085EAFF is the end of block 2. I have no idea what 00000100 or 0001E8C0 could be.

looking at this does the checksum really include 40240?

Also, there is apparently a set of bytes to enable/disable the checksum calc. it's at 400EC. I'm not saying that disabling the checksum is a great idea, but if you don't have the ability to correct it yet, it might be useful for testing purposes. I'm betting it will be disabled if you change those bytes to anything other than 0 (perhaps FFFF). It would be easy to test, set them to FFFF and change the checksum to the wrong value.