[device7]

Situation:

Ps3 Sem-001 mobo (CECHG)
Dumped both nands with ~8 bad blocks
Using Win7 64bit + Winskeet software + 25cm usb cable + Progskeet 1.2 + Adaptor + NAND Clip + strong CLAMP + always renew contacts with Defluxer

For 3.3V and GND used wires from OLD PC CASE (wire which was from "turn on" button to mobo)

Every time i get different bad blocks. Sometimes i get only 2-3 sometimes more. Bad blocks i get for example in that order:
2,5,17,53,114,505. all other ok. Very often the second bad but one time it was correct.

1. I can remap bad blocks but - if it didnt read correctly how could it flash patched nands back correctly?
2. Installed Serial from github and downladed nandway.py. Try to run nandway.py and get error

Traceback (most recent call last):
File "C:\winskeet\dumpai\nand0\nandway.py", line 16, in <module>
import serial, time, datetime, sys, struct
ImportError: No module named serial

3. Every read i get different BAD blocks. Do you think remapping can help?

[Sarah1331]

No use your progskeet power to power NAND nothing else

[device7]

Originally Posted by Sarah1331 No use your progskeet power to power NAND nothing else

I use power from progskeet to pcb adaptor

[Sarah1331]

Your clip is not connected correct

[Abkarino]

Originally Posted by Sarah1331 Not got one to hand try explane if you want (please)

You can just make your own RSOD just to test.
Just Dump your NOR flash change any bytes in cVTRM area then flash it back, then you will get the RSOD.

I had fixed a dump for 3absiso and i'm waiting his feedback about this if it fixed his friend RSOD or not?

[zoom33]

Originally Posted by device7 Situation: Ps3 Sem-001 mobo (CECHG) Dumped both nands with ~8 bad blocks Using Win7 64bit + Winskeet software + 25cm usb cable + Progskeet 1.2 + Adaptor + NAND Clip + strong CLAMP + always renew contacts with Defluxer For 3.3V and GND used wires from OLD PC CASE (wire which was from "turn on" button to mobo) Every time i get different bad blocks. Sometimes i get only 2-3 sometimes more. Bad blocks i get for example in that order: 2,5,17,53,114,505. all other ok. Very often the second bad but one time it was correct. 1. I can remap bad blocks but - if it didnt read correctly how could it flash patched nands back correctly? 2. Installed Serial from github and downladed nandway.py. Try to run nandway.py and get error Traceback (most recent call last): File "C:\winskeet\dumpai\nand0\nandway.py", line 16, in <module> import serial, time, datetime, sys, struct ImportError: No module named serial 3. Every read i get different BAD blocks. Do you think remapping can help?

From my experience you're close to get a brick if you can't make sure you have a good dump.
Trust me, never but really never flash on a NAND that is not 100% safe.
You're good to go for a NAND desoldering and use of a pcb with an adapter (ebay is your friend).
I don't give a **** for that NAND clip !!! (i know very good because i have one).

[Sarah1331]

I sell the perfect adaptor for this

Check the link in my sig

[zoom33]

Originally Posted by Effleurage top and bottom should be treated separately. this is not related to whether you patched the NAND or not. you can call this script - "NAND copy". it overcomes the problem of bad blocks which are found (during manufacture or later) in the NAND. What the script does is: 1. strip all the bad blocks from src_nand. (again - it can be either patched or not). 2. find all the bad blocks in tgt_nand - that is why i need new_nand.bin 3. remap data we have in (1) to the tgt_nand using what we found in (2). ************* [ - Post Merged - ] ************* ok, first of all, I fixed the script - so you can take it from the hjudges/NORway repository and retry. what you did seems right to me (all the calculations). the only thing that is a) in case you have several bad blocks, and b)missing is the remapping to the new NAND. If the new NAND doesn't have any bad blocks - you are fine. otherwise you have to remap

Nandway.py is working now, but still doesn't do the job of remaping like it should.
I have tested it and it mistaken a bad block with a good block.

The only way to make sure a block is really bad is by trying to overwrite every byte within the NAND it self or as stated by sarah1331 you have to check every single byte has a $00 value.

Anyway flashing the dump back with one or two remapped bad block doesn't work.
If you use flowrebuilder 4.2.3.0 (last version) it creates 4 txt files.
those files are saying there is a bad maping within the dump.
************* [ - Post Merged - ] *************

Originally Posted by Sarah1331 I sell the perfect adaptor for this Check the link in my sig

even better !!!
choice is yours "device7" ... invest in something you can rely or make a brick PS3.

[Sarah1331]

Also you only need to remap if florebuilder says errors it can handle most bad blocks just not all

[device7]

I have also tried to dump NANDS from different Cok-002 boards today.

Results:

1. Two different boards and the same results with all NANDS - Winskeet identifies "Unknown Device 0000000000" but i set BigBlock, 64, 1024 and dumps without bad blocks. Of course i dump just for testing. FLow rebuilders tells me that both NANDS have 1023 bad blocks.
2. Until i dont get all ok i would not flash.

I have also tried:
1. Millions of cleanings flex cables, clip, chip.
2. Finally i connected 3.3V with 5cm short cable.
3. Tried various miniUSB cables from short 5cm, to long. Different qualities.
4. Tried various flex cables.
5. Tried computer with Pentium M 1,7GHz and XP (old one) and eSata (new one).

THE SAME!!!!!!!!!!!

Bth, i use Progskeet 1.2. Tomorrow i will try to update bitstreams and also try to dump more different boards and maybe linux...
also i will try to supplye 3.3v from different source to adaptor and also maybe to Ps3 mobo (now mobo is not powered)

I would buy that Sarah pcb but all the procedure should look like this, am i correct?
a. Desolder chip0
b. Solder on PCB
c. Dump
d. Desolder from PCB
e. Desolder chip1
f. Solder on PCB
g. Dump
h.Unify and unified dump
g. Split
h. Flash on chip1
g. Desolder chip1 and solder chip0
h. FLash on chip0
g. Desolder chip0 and solder both chips on mobo

Complicated and long maybe 2,5 hours procedure as with clip it should be maybe 45-60min

But, of course,if i not succeed with clip i buy 2pcs of that PCB from Sarah. What price would be for 2pcs including shipping to Lithuania? I can pay by paypal or CC. Direct payment cost 10pounds and plus some for you so it is nonsence for small payments.

Bth, i have done 55pcs of NOR ps3 with E3 and clip - everything just FINE!