nand keeps corrupting

[afsilver]

decided to RGH my jasper SB, soldered pin-headers to the mobo and read the nand with one bad block
"Error: 250 reading block 2B7"
6times!
put it back in it's casings and booted straight to RRoD!
read back nand once more and compared with nand1.bin
got :
1024 block(s) in nand image
832 non-matching block(s) found

so i flashed nand1.bin and booted 2 red lights.
removed and reapplied thermal paste. booted RRoD!

compared nand got one non-matching block "3DE" i think

wrote nand1.bin back and booted to dash, was just about to connect controller and there it goes again, RRoD!
1024 block(s) in nand image
832 non-matching block(s) found :facepalm:

what is happening? no matter what i do it keeps giving non matching blocks and RRoD!

edit: i actully don't know if it is a BB or SB :facepalm::facepalm::facepalm::facepalm:

that's the reason i never flashed xell or anything cause i wanted to check... :rolleyes2:

 

[splatem]

Xnandhealer will tell you if it is a big or small block Jasper.

 

[afsilver]

if i load nand1.bin in Xnandhealer it says smallblock but maybye because i only dumped too 3FF?

ive set nandpro to read all the way to 7FFF, and it's currently @B61 with 3 bad blocks:
2B7
6B7
AB7

since it keeps reading i guess i do not have SB?

---------- Post added at 21:53 ---------- Previous post was at 21:51 ----------

but should it keep corrupting just because i only read too 3FF? makes no sense too me..

 

[ishockin]

Pics of pin headers?

 

[afsilver]

ah sh*te! if ive managed to short it out i'm fckd rite? i'll get som pics up right away!

 

[afsilver]

xnandhealer confirmed jasper SB

tested with voltmeter can't find any short's

 

[splatem]

It is a bit hard to see in the photos, but the solder doesn't seem to have flowed very well on the pins under j1d2. Perhaps the connections aren't as good as they could be there, have you checked them with a magnifier?

 

[afsilver]

Would'nt that give me bad reads/writes? Shoulden't make a perfect write unreadable after a good boot.. Can't make any sense of these...

If i write nand1.bin and read back and compare before boot, it matches 100%. So problem is during boot/use..

Sent from my GT-I9100 using Tapatalk

 

[afsilver]

So i guess i have to write good nand, then desolder and clean.
Previous experience and all you guys telling me it's the soldering, unless some genius comes along with another explanation..

If ive managed to f-up the nand points how am i going to manage to install the coolrunner

Sent from my GT-I9100 using Tapatalk

 

[diaboliq20]

if you boot between reads, they wont match
lols - ~I don't think you've fudged the nand points.

 

[afsilver]

Yeah i would expect some diffrence but 800 out of 1000 non matching? and all the files show up as wierd signs (#%¢§) instead of for example xenon.elf and similar.. So the nand image is corrupted after one boot. Will post a screen shot or my corrupted nand image once i get back from work

Sent from my GT-I9100 using Tapatalk

 

[esbmaepo]

Yeah i would expect some diffrence but 800 out of 1000 non matching? and all the files show up as wierd signs (#%¢§) instead of for example xenon.elf and similar.. So the nand image is corrupted after one boot. Will post a screen shot or my corrupted nand image once i get back from work

Sent from my GT-I9100 using Tapatalk

You could try ditching nand-x and solder wires directly to points. You never now....:rolleyes2:

 

[afsilver]

you mean remove pin-headers? i'm tempted but i have no spare cables for the nand-x, so if i want to solder the wires directly i have to cut the connectors of the cables.. i want to try abit more before i cut them off..

 

[afsilver]

here you can see the diffrence before and after a boot

 

[DreamZ]

crc wont match if you read, boot console and re read, read a few times make sure they ALL match.

 

[afsilver]

i did read the nand 6 times and they all match! i'm aware that nand wont match 100% between boots but over 15files missing and all of the remaining files has been badly renamed.. you can't say that's normal?

just redid j1d2 as they were abit poorly done as splatem suggested, got some proper lighting and a magnifier!

 

[esbmaepo]

you mean remove pin-headers? i'm tempted but i have no spare cables for the nand-x, so if i want to solder the wires directly i have to cut the connectors of the cables.. i want to try abit more before i cut them off..

I understand. You could build a LPT cable (if you haven't got one) then you would have an extra cable

 

[afsilver]

booted 5 times straight too RRoD, then once with two lights and fans spinning up for 1 sec. then twice too dash (never long enough to reach system info) last 10 attempts was only RRoD some straight away others during startup animation, if it is relevant..

gonna try donor files see if anything changes..

I understand. You could build a LPT cable (if you haven't got one) then you would have an extra cable

will have to look into that, neve tried before and components are hard to come by in short notice, most likley have to order them.. thanx for tip!

 

[Martin C]

I understand. You could build a LPT cable (if you haven't got one) then you would have an extra cable

Why the hell would he want an LPT cable if he's dumping via USB?

Confusion here. Please explain!

---------- Post added at 16:27 ---------- Previous post was at 16:27 ----------

OP - what bad blocks do you get in NANDpro and 360 Flash Dump Tool?

 

[afsilver]

-> Block ID 0x02B7 found @ 0x3FF [offset: 0x0107BE00]

but i got a new issue now! god dammit! when i try to read my nand i get:
Incompatible ARM hardware for this version of NandPro!
Can not continue

nand-x is version 3.0 and i am ussing nandpro 3.0 as i have the last 2 days...

if i unplugg and plugg back the usb it starts reading nand but with a bunch of errors, mostly error 0

this is after i foolishly tried the repair bad nand option in jtag tool.

downloaded clean jasper 16mb nand from jtag tool and put my nand1.bin in as "bad nand", flashed the resulting image to nand gave RRoD and the above errors when attempting to read nand..