Electronics > Repair
Agilent 34461A corrupted flash
<< < (23/41) > >>
analogRF:
by the way the areas that I wrote in the NAND (starting from 0xE0000) look exactly identical to the areas I copied from
including the oob data


so I dont think writing back into the nand is gonna help, is it?
ElectronMan:

--- Quote from: analogRF on July 30, 2023, 10:23:49 pm ---still I am pretty sure the whole 128MB of NAND is not bad and I am not even sure the 0xC0000 area is bad either.

perhaps there is still a problem with the pboot itself or it is reading its parameters from somewhere else , not from 0xc0000

--- End quote ---

I've considered a possible issue with PBOOT, but the problem is that it has a couple of different levels of checksums protecting it. There is a table inside it that it seems to use for the check. If that got corrupted, it could cause the appearance of a failure that is not there.

But that is extremely unlikely given that the checksum that UBOOT sees when you list images shows Ok, and the gzipped nb0 for PBOOT has its own CRC.

It is more likely something in the spare area isn't matching up. A hardware issue could cause a bit to appear stuck, for example.

The hope is that by rewriting you can force a re-write of the spare by the same mechanism that does the checking (FMD_Write). If that doesn't correct the issue, then it is looking more like a flash hardware problem.
ElectronMan:

--- Quote from: analogRF on July 30, 2023, 10:29:45 pm ---by the way the areas that I wrote in the NAND (starting from 0xE0000) look exactly identical to the areas I copied from
including the oob data


so I dont think writing back into the nand is gonna help, is it?

--- End quote ---

Did you try rebooting since? You can test the theory, as it should check the block you wrote (it was checking it previously) when the first one fails.

If it still shows an error on that block (assuming you wrote the entire block) then it is not looking like an error that is correctable by re-writing it.

And erase cycle on that block might help, but now we're getting into more dangerous stuff.
analogRF:

--- Quote from: ElectronMan on July 30, 2023, 10:32:28 pm ---
--- Quote from: analogRF on July 30, 2023, 10:29:45 pm ---by the way the areas that I wrote in the NAND (starting from 0xE0000) look exactly identical to the areas I copied from
including the oob data


so I dont think writing back into the nand is gonna help, is it?

--- End quote ---

Did you try rebooting since? You can test the theory, as it should check the block you wrote (it was checking it previously) when the first one fails.

If it still shows an error on that block (assuming you wrote the entire block) then it is not looking like an error that is correctable by re-writing it.

And erase cycle on that block might help, but now we're getting into more dangerous stuff.

--- End quote ---

yes I mentioned that I rebooted and I still get the same bad block error at sector 0x1c0 which is where I copied into. no change
also the oob data in those areas that I wrote into now are exact copy of the ones I copied from.

I just opened up my good 34461A and dumped the same area of the nand. The oob data looks exactly identical to what i see on my bad meter
so I am 100% sure it is not about those OOB data

ElectronMan:

--- Quote from: analogRF on July 30, 2023, 11:01:16 pm ---
--- Quote from: ElectronMan on July 30, 2023, 10:32:28 pm ---
--- Quote from: analogRF on July 30, 2023, 10:29:45 pm ---by the way the areas that I wrote in the NAND (starting from 0xE0000) look exactly identical to the areas I copied from
including the oob data


so I dont think writing back into the nand is gonna help, is it?

--- End quote ---

Did you try rebooting since? You can test the theory, as it should check the block you wrote (it was checking it previously) when the first one fails.

If it still shows an error on that block (assuming you wrote the entire block) then it is not looking like an error that is correctable by re-writing it.

And erase cycle on that block might help, but now we're getting into more dangerous stuff.

--- End quote ---

yes I mentioned that I rebooted and I still get the same bad block error at sector 0x1c0 which is where I copied into. no change
also the oob data in those areas that I wrote into now are exact copy of the ones I copied from.

I just opened up my good 34461A and dumped the same area of the nand. The oob data looks exactly identical to what i see on my bad meter
so I am 100% sure it is not about those OOB data


--- End quote ---

I am still looking at the code to see exactly what it is verifying in the spare area. If I figure out what it is looking for, maybe we will have some idea of why it is rejecting it.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod