Hitachi HDT721010SLA360 not spinning - TVSs OK

[fzabkar]

I don't know if it applies to your HDD, but in some models the MCU controls the -5V supply by sending a command to the motor controller. The presence of this supply would suggest that the MCU and motor controller are communicating with each other. You may be able to confirm this by preventing the MCU from waking up and watching the -5V test point. Often grabbing the crystal (Y1) between your fingers is enough to prevent it from oscillating, so I would do this while you are powering up the drive. If the test point remains at 0V, then this would suggest that the MCU is working. Otherwise, if the voltage remains at -5V, then the result would be inconclusive.

[int0x13]

Some news:
I tried to block the crystal which is the black rectangle component marked Y1 right? It didn't work at all either with finger on it or with putting the only pin around to ground (0.5v on the only filled in hole on the left of the rectangle). Either way the L1 inductor is always at -5v and all the voltage seem present as well as vcore on MCU so its starting anyway. I've seen the smd crystal is mounted on 4 legs on its 4 angles I don't know if those are pins?
I've unsoldered the nvram chip(harder than I thought to get it) and the HDD is starting the same ( all voltages, no spinning, bios dont detect it).
Way I do not understand why its not detected by bios even if its not spinning (which prevent the MCU to load the whole few on platter) shouldn't HDD PCB communicate with bios to detect and display the HDD model?
Now I will try to read the nvram chip I've unsoldered.

[fzabkar]

Attached is a sample NVRAM dump from an HDT721010SLA360.

Attachments

NVRAM_SLA3_E2PRST3NA3AA_944ACF71.zip

(1.27 KiB) Downloaded 367 times

[int0x13]

It confirms me that something is wrong with my dumps. I will try to dig it more to understand where is the bug. But i find it strange because there is a chunk of memory with zeroes around 0x200 which i have in my dumps. So i suppose there is a slight issue in the way my device is interpreting the bytes. Thank you for helping fzabkar. By the ways ive ordered a new PCB already, we will see, but i really hope my NVRAM chip is not corrupted.

[fzabkar]

I would read the donor's NVRAM to confirm whether your methodology is correct.

[fzabkar]

ISTM that the contents of the suspect NVRAM have been rotated to the left by 6 bytes and then mangled.

[HaQue]

Or possibly read with wrong chip parameters?

[fzabkar]

The OP has provided the NVRAM dump to me via PM. It appears that the contents have been byte swapped, rotated right by 2 bits, and then rotated left by 6 bytes.

[int0x13]

Just to close my case, my Hitachi 1TB HDD is 100% fixed after changing the PCB and swapping NVRAM.

[HaQue]

Just for clarification, The NVRAM that worked was the one where dump that didn't look right?

[fzabkar]

The programmer settings for the original dump were incorrect.

The chip reader was sending an 8-bit address whereas the chip was expecting 10 bits. That accounted for the 2-bit right shift. Also, the first [16-bit] address was misinterpreted as 0b11 rather than 0, so this was responsible for the 6-byte left shift.

After amending the script for 10-bit addressing, the dump was correct, and its checksum verified.

http://www.alexsoft.org/viewtopic.php?t=80&p=4590#p4590

[HaQue]

Thanks Franc, so I was right

I wondered if there was some tricks to try and thwart playing with the firmware..

[int0x13]

Just for clarification, The NVRAM that worked was the one where dump that didn't look right?

Yes, the original NVRAM worked and has always worked, it was just a problem with my dumper. With the help of fzabkar we managed to understand and fix the "corrupted" dump. I want to thank him for his help to understand, as it was a little tricky. I wanted to check the terminal, because i've never seen on Hitachi drive, i've never seen any terminal but seagate one anyway. But ive never found the good pinout and/or serial settings combination.

I wondered if there was some tricks to try and thwart playing with the firmware..

What do you mean?

[fzabkar]

I think HaQue is saying that at the outset it wasn't clear whether the corrupt NVRAM dump was due to a problem with the reader, or whether there was some kind of "protection", either in the firmware or in the chip itself.