USB SPI Flasher with PIC18F2455 - now with source and binaries

[rolf2]

how about resistors on 1 , 13 , 26 pins ?

[Arakon]

13? What resistor is supposed to be there?

Looking for usb interface device
FlashConfig:01198010
Starting Block:0x000000
Ending   Block:0x0003FF
Reading
Error:  A3 reading block 0
Error:  AC reading block 1
Error:  B9 reading block 2

The oddest thing is, the flashconfig is alright.

edit: pin 26.. should that go directly to GND since I don't need an external programmer anymore (chip is already programmed and all)?
edit2: makes no difference. I've gotten up to 1B for reading before it starts erroring, but it ALWAYS eventually goes nuts. I tried on win7 with the updated driver, on winxp with the older driver and the updated driver, none of it makes any difference.

[nexus148]

My card dont detect the flash. I have used nandpro 2.0a and b.
I use W7, in lpt version work ok but slow.

Code:

C:\nandprob>NandPro.exe usb: -r16 my.bin
NandPro v2.0b by Tiros

Looking for usb interface device
Could not detect a flash controller!
Can not continue

Any help?

[rolf2]

if you are using usb as power for pic, then ya, you dont need any resistor on 13, but you connected this pin to +5 V right ? I got similar problems when pic sometimes is detectet and after some time not when pin 13 is left floating ...

[dak]

edit: wtf.. I found that pin 13 had a bad connection.. fixed it, started dumping, after 40 or so correctly dumped blocks it disconnects again and then only errors again too.

Finally got mine to work under Win XP, but now im facing the same Problem.
You use PIC 18F2550 ?

[Arakon]

13 is connected to 5V, yes. I use a 18F2455.

It seems the error always happens at 1B for me now.. slows down and then errors starting at 1B, with totally random error codes.

[rolf2]

hmm, then maybe there's to small capasitor on vusb ?

[Arakon]

220n.
But I believe I found the issue, thanks to [c0z] from #freexex.. I flashed the older 3a firmware to the PIC and now it's currently already past half of the dump.

[dak]

220n.
But I believe I found the issue, thanks to [c0z] from #freexex.. I flashed the older 3a firmware to the PIC and now it's currently already past half of the dump.

2000 and still going (256MB Nand), with 3b plus it stopped at ~1000. Thanks Arakon

[nicomic]

Hello eveybody,
like i said before i try to flash my pic18f2550 but it doesnt work
i found that the parralel port that i made wasn't good because the schematic wasn't the same as for the free60 because the pin we see to solder aren't the same.
so now the hardware is detected in winpic but not the pic it says unknown.
i put a 5V alimentation to make sure the voltage is good but it doesn"t work, can sombedy know whyµ?
thank you and sorry again for my poor english

[cory1492]

220n.
But I believe I found the issue, thanks to [c0z] from #freexex.. I flashed the older 3a firmware to the PIC and now it's currently already past half of the dump.

2000 and still going (256MB Nand), with 3b plus it stopped at ~1000. Thanks Arakon

This is just an issue with dumping then? I'll have a look at it - nothing I did should affect dumping 

[Arakon]

Not sure, I didn't actually try to write yet. But for me, it ended up always messing up at 01B, with both nandpro 2.0a and 2.0b.. with the older hex, 2.0b doesn't work, but I got four matching dumps easily with 2.0a.

[dak]

Sorry but cant tell you if it always happend at the same block, but it was always at ~1000. Had to reconnect USB then.
Didnt try writing yet.
Components used were exactly those from the first post from reichelt, but PIC 18F2550.

Edit: I got 1 good dumb but now it crashes again.

how about resistors on 1 , 13 , 26 pins ?

Id have 3x100n and 3x220n here. Should i try 220n on all 3 pins ?

[cory1492]

Those of you with issues (especially if dropping usb in windows), using resonator or crystal? Resonators are usually labeled for +/- 20% off their rated value... also the cap value used on VUSB? See if there is something common going on hopefully.

I begged my chip to give me the issue with my falcon this morning for a good hour so I could debug it, but it wouldn't do it for me - worked perfectly on every dump 

[fatjoe]

Finally i got this working! Dumped nand twice with zero errors.. How do i check that nand dump is valid? I compared dumps and those are identical.. Is it enough? Big thanks to all who have helped me!

[dak]

Those of you with issues (especially if dropping usb in windows), using resonator or crystal? Resonators are usually labeled for +/- 20% off their rated value... also the cap value used on VUSB? See if there is something common going on hopefully.

1x CST 12,00 - 12 MHz, +-0,4%
1x Z5U-2,5 220N, 20%, Voltage: 50-100V, on VSUB
1x Z5U-2,5 100N, 20%, Voltage: 50-100V, on VDD

Ive built a second flasher with new components, this one stops at ~2000.

Which caps should i use instead of those ?

[joeyddr]

As Sandungas mentioned on IRC and from my own experience i found that the trace lines on the breadboard should be thick especially the 5v and ground.... also the USB data lines i wound up soldering them directly to the legs of the chip and then i was able to get a good read without disconnect.

[Arakon]

I just rebuilt the programmer using a DIP chip.. and guess what? Instantly works, using nandpro 2.0b and the plus2 hex. I am quite sure now that the issue was the resonator, because I used a crystal this time (8 mhz cause I had that sitting around), and it works perfectly.


edit: I'm getting 2-4 differences between dumps, always in about the same area (around 42E384, filled with C0, 00, E0 and FF). Could this be related to a bad block (103)? The odd thing is, with the flawed flasher and older version, this didn't happen, the dumps matched.
The oddest thing is, even with infectus I get mismatched dumps, also in that area.

[maximilian0017]

I'm getting 2-4 differences between dumps, always in about the same area (around 42E384, filled with C0, 00, E0 and FF). Could this be related to a bad block (103)? The odd thing is, with the flawed flasher and older version, this didn't happen, the dumps matched.
The oddest thing is, even with infectus I get mismatched dumps, also in that area.

What do you mean with around 42e384?
You would think this would happen around 40c000 to 40ffff for block 103(out of my head)
The xbox wasn't powered on in between reads?

Using a 12 Mhz crystal and a 47mF for vusb and short wires i have had 0 bad R/W out of about 40 R/W cycles, using one of the older flashes, will try newer one as soon as the next shipment of PIC's are delivered :-)

[Arakon]

around there as in usually some bytes above or below, but the differences are totally random in that area.
the console wasn't powered. I took 3 dumps with infectus and got the exact same behavior. I flashed back an old 1888 dump from when I downgraded the console, and flashed back one of the dumps I had made with the USB programmer, and both appear to work flawlessly.

edit: I just compared to one of the default XBR versions, and that area has data in it there. So I have to assume that the differences appear indeed in the bad block, and differ due to that.
edit again: bingo.. the data that is in the "bad area" in the default XBR can be found at the end of the dump in my dumps.. so as I assumed, that is a bad block and randomly produces different data (within the bad block only).