USB SPI Flasher with PIC18F2455 - now with source and binaries

[Grim187]

4.40 from the diodes, pin 28 is 0.53.

50k (my thumb, dont ask) between pin 1 and pin 28 got it to program and it verified %100 w/o the 50k, it dose look a little different with a hex compare.
http://www.megaupload.com/?d=XBOBFCKO

[maximilian0017]

4.40 from the diodes, pin 28 is 0.53.

50k (my thumb, dont ask) between pin 1 and pin 28 got it to program and it verified %100 w/o the 50k, it dose look a little different with a hex compare.
http://www.megaupload.com/?d=XBOBFCKO

Hmm strange, i flashed again with v3 and also read back the flash and i also have small differences if i look at the original and the read back version, maby a different program software generates a different file?.

Can you read the original hex into WinPic800 and then your dump?, please look at the fuses for both.

Have to build the actual second board to try to connect to it through usb, that will give me 100% assurance that the flash went well over here.

Did you change the fuses before flashing?

[DarkstarTM]

Hmm strange, i flashed again with v3 and also read back the flash and i also have small differences if i look at the original and the read back version, maby a different program software generates a different file?.

Where exactly are the differences and what exactly differs?

[maximilian0017]

Where exactly are the differences and what exactly differs?

Usb works like a charm, bootloader was picked up directly, so no need to worry.

I deleted the old hex files, if you think it is important i can remove the chip and reprogram it.

You could also check Grim187 his file with the 1.2 hex

[maximilian0017]

Ok, I retrieved the correct files from the garbage bin i hope
If i visually(not 100% sure ofcourse) compare the code/data/settings in WinPic800 they are the same.
If i compare the HEX files with Hexworkshop they differ in a lot of places.

So it does look like WinPic writes the hex file differently, but as it works over here it is nothing to worry about A.T.M.

Maby i can get the hearders and the wiring done tomorrow and test this baby out!!

[maximilian0017]

Is it possible that in the v3 release the PicXBoot_12.hex is still the old one?

With:
RB5 -- 10k -- Ground
C6 -- Jumper -- Ground

it will not enter the bootloader but just executes the main PicFlash.hex(witch works like a charm reading the NAND, maby tomorrow i have the time for flashing)

For good measure i used a DMM to check the connections and they are correct.

[fail]

I failed to get it work ! 

I' m stuck after programming the bootloader picflash hex via USB...

I type NandPro.exe usb: -r16 nand.bin in dos prompt and it says No usb device detected !

I didn't  connect the flasher to the motherboard because i wanted to test it first.

Winpic800 Programming is ok, USB programming via Bootflasher is ok (strange thing is that if i try to use the erase (only) function it returns an error), flasher however is not recognized by Windows (of course after removing the ground bridge), it doesn't ask me to install the drivers.

I'm using 12 MHz resonator (without 22pF capacitors to ground because i'm lazy)...

Need help here !

[cory1492]

Have you wired RC2 to 5V? It's being used as USB Sense, without it windows will not detect the device (or will only erratically detect it) for the libusb driver while the bootloader is set up to always think usb is sensed. Since USB programming detected and worked I don't think the resonator is your issue, in my limited understanding: the caps are used to kick off/prime a crystal oscillating and aren't really needed for a resonator.

[fail]

Usb Sense is connected.

I keep failing, indeed it's my name 

PICDEM FS USB gives me  the following output when i try Erase Device:

WARNING - Failed to Verify.
WARNING - Actual Value :00
WARNING - Expected Value at Address 0x005800 : 0xFF
MESSAGE - Verifying Memory starting from 0x800 to 0x7FFF...
MESSAGE - Erasing Flash starting from 0x800 to 0x7FFF...

It gives me this when i use  Program Device:

MESSAGE - Programming FLASH Completed
MESSAGE - Erasing and Programming FLASH...

Addr.  00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F

I'm banging my head against a thick wall XD

[cory1492]

Used winpic800 to program the current bootloader hex from the OP with the config modification that LVP is disabled and clock set to 24MHz (which is what I'm using.) No issues with the bootloader (aside from reset stuff mentioned below), and the programming exe programs and erases without issue/error. Execute command does not seem to work (WARNING - Failed to RESET board). Replug with RC6 floating and the proper libusb device is detected by windows. Perhaps there is an issue with your PIC's flash? Does the error occur always at 0x5800? The xflash program itself should end before 0x5800, it's not all that big of a program which should explain why it flashes without issue. (edit:/ also, be sure to pull pin 1 to power with 10k.)

maximilian0017: I can't get the current bootloader to enter bootloader program mode myself via RC6, trying to get to update mode using the python script results in "usb.USBError: usb_control_msg: sending control message failed, win error: A device attached to the system is not functioning." (I get an expected timeout error trying to read a block while it's not connected to a 360.) (edit:/rebuilding bootloader from current git source has reset from the exe working, but not from the python script; as well, RC6 pulled to ground now works to go back to programming mode so I presume the current hex is indeed not the correct ones, here is my compile of what is currently in git if you want to tinker until DarkstarTM updates and it is working with the py reset and RC6 pins without issue.)

[maximilian0017]

here is my compile of what is currently in git if you want to tinker until DarkstarTM updates and it is working with the py reset and RC6 pins without issue.)

Ah thank you, thought i was going mad 

Reading/flashing was a breeze, worked the 1st time, superb work.
(beat that Olimex LOL)

DarkstarTM:
Are you planning a version that also takes over the Southbridge/Jtag hack part?, this would be nice for all the people who have a different MB version.
Maby a smaller pic could do this?, or maby an ATTiny or something.

[DarkstarTM]

the caps are used to kick off/prime a crystal oscillating and aren't really needed for a resonator.

Yes they are needed. But in a three-pin-resonator, the caps are included.

[Grim187]

i built up the flasher on a breadboard and got it programmed and now when i plug it in it try's to install "MemoryAccess" and cant find a driver, the one included in "PicXFlash_V3" dosent work.

http://i12.photobucket.com/albums/a246/Grim187/MemoryAccess.jpg

[cory1492]

You need to use the libusb driver from NandPro (if under x64 you also need to reconstruct the inf and deal with test signing.) The driver you are trying is the microchip driver for usb programming mode.

edit:/ here is just the driver, though I rebuilt the inf to default with the libusb inf wizard tool so it should be OK for x64 (I have test signed a set of this with DSEO and it works fine with nandpro and the xflasher python script on win7 x64 with this PIC device - though you do need to track down and add DLPortIO.DLL beside the nandpro exe for it to run even if not using parallel port.)

[fail]

Could someone with a working board use the erase function from PICDEM FS USB ?

I would like to know if the error it gives me is not normal ...

Other question ... Does the board need to be connected to the xbox to be recognized by the PC ?

[Grim187]

i got the driver installed but nandpro cant find it

nandpro usb; -r16 nand.bin
NandPro v2.0 by Tiros

Testing LPT device address:0378
Testing LPT device address:0278
Testing LPT device address:03BC
Could not detect a flash controller
Fatal Error

nvm, i used nandpro usb: -r16 nand.bin and it worked.

it will only read up to 0050 then i get errors like 21c, 210, 214 and 218
i just wrote 128 blocks (xell update) and that worked w/o error.
after writing i tried reading agian and got to 167 before i got an error.

[Grim187]

Could someone with a working board use the erase function from PICDEM FS USB ?

I would like to know if the error it gives me is not normal ...

i got the same error erasing, it programed fine so i assumed that it worked.

[cory1492]

Could someone with a working board use the erase function from PICDEM FS USB ?

I would like to know if the error it gives me is not normal ...

Other question ... Does the board need to be connected to the xbox to be recognized by the PC ?

It erases fine here - but the PIC I am using is what the code was originally designed for (18F4550) so perhaps that is not much help. No, it doesn't need to be connected to the xbox to be recognized by the PC in either bootloader or picflash modes (so long as you are using USB for 5V/power at any rate.)

Grim187: I had similar problems with random disconnects/not detected/sequential error numbers because I had forgotten to pull pin 1 up with a 10K resistor to 5V. When you get such errors, is you PIC disconnecting from USB? Which code are you using? If it's the hex I uploaded the other day, you should check your diagram against the revised one I posted/edited into my post a couple pages back (esp. for the pull down resistor on RB5 if you have left LVP enabled.)

[Grim187]

im using the v3, rb5 is grounded, pin 1 > 1k > +5, rc6 is nc, i will pull up device manager and watch it next time i try a read but i dont think it disconnects.
ive checked everything it could be 5 times now, it gave me a error after 50 3 times in a row and after i programed the xell update (iirc the update was finished at 167) it now gives the error after 167, so i think i might be able to get the reading working if i can program to 3ff (just a theory).

[maximilian0017]

It erases fine here - but the PIC I am using is what the code was originally designed for (18F4550) so perhaps that is not much help.

Reading/writing works great over here with the original pic & the v3 archive, no problem there.

Things i found to be strange are:

1. Shutting down the pc with a booted xbox360 will turn off/crash?!? the 360.(happened at least once)
2. Connecting the pc while the 360 is connected to the mains will result in the pc not finding the usb powered pic.
First connection the usb and then connection the 360 to power solves this. (happened a couple of times)
3. Getting into the bootloader through the jumper does not work(cory1492 stated that this is the old bootloader, new one solves this)

Art 2003 programmer/Winpic800:
1. Sometimes when writing will not work correctly, just reset the pc and everything is fine.
2. When reading the PIC(with WinPic800) after flashing the Bootloader(verify 100%) this results in a different HEX than the original, it seems that WinPic saves the Hex a different way, conclusion: if WinPic verify's with 100% the Bootloader is correctly written.

Documentation:
A reference to J1D2.6 or J2B1.12 would be nice for the VSS pin.

Overall:
Was certanly worth it, works like a charm at a fraction of the cost of the Olimex board.
Flash/read in about 6 minutes, flawless