Helium Filled Drives

[ddrecovery]

Anyone done much research on these drives in terms of head swaps?
Obviously we have several issues such as the sealed lid, helium, donor costs, locked FW etc etc.
I have a WD100 with stiction and am thinking of buying some donors to experiment with to see of head swap is actually possible.
Any information would be appreciated, and my findings will be shared.

[melvin]

Didn't find a way to open He drives without leaving a mess inside. Besides they will not work without He.

[ddrecovery]

Didn't find a way to open He drives without leaving a mess inside. Besides they will not work without He.

I have read reports that drives can be recovered without He. All depends on how much data the client has as the heads die after a while. There 'reports' could of course be covered in the delicate smell of bullshit.....

[fzabkar]

ISTR one thread where it was claimed that some models won't spin if the helium sensor reports a low He level.

Even if a head swap is achievable physically, I wonder whether the donor's head adaptives would need to be transferred to the patient's ROM. I have written a tool to extract these adaptives from HGST ROMs, but I haven't been able to work out the CRC algorithm. I know that some technical information has leaked into the DR community, but I haven't seen it.

[northwind]

We've done some tests, they all failed

Even if you manage to cut open the lid (we used laser) without making a mess, and solve the He problem (we didn't, we just wanted to see if the drive will spin up and calibrate after a head change) and you unlock the FW (which, to the best of my knowledge is not viable at this moment), you need to make the new heads boot up. And, at least to our test, they didn't.
Most probably due to the adaptives Frank is referring to.
The drive DID spin up without He, though.

[ddrecovery]

Thanks for the comments. I will report my findings as I progress.

[fzabkar]

We've done some tests, they all failed :(

Even if you manage to cut open the lid (we used laser) without making a mess, and solve the He problem (we didn't, we just wanted to see if the drive will spin up and calibrate after a head change) and you unlock the FW (which, to the best of my knowledge is not viable at this moment), you need to make the new heads boot up. And, at least to our test, they didn't.
Most probably due to the adaptives Frank is referring to.
The drive DID spin up without He, though.

What if you replace the heads and PCB as a set, rather than transplanting the adaptives?

[ddrecovery]

What if you replace the heads and PCB as a set, rather than transplanting the adaptives?

I guess that might work if the patient SA and PCB are compatible. Impossible to know on these drives at the moment.

[jono-ats]

What if you replace the heads and PCB as a set, rather than transplanting the adaptives?

Not likely. I would imagine there are some platter adaptives that have to agree with the ROM. For example, in older HGST products, the NV-RAM (within ROM) points to a boot location on the media.

[saltwater]

Without helium i think the air makes more turbulence and pressure than helium, and the heads can't fly on the calculated height. That's why HSA can't calibrate.

[fzabkar]

I would imagine there are some platter adaptives that have to agree with the ROM. For example, in older HGST products, the NV-RAM (within ROM) points to a boot location on the media.

The only adaptive section that looks like it might point to SA cylinders (or zones ?) is the "CYL" component in the FSVS module. Otherwise there is nothing which looks like the old style "NVRAM". Of course ICBW.

Adaptive sections in FSVS module in HUH728080ALE604 ROM

Code:

size   ID    Description
-----  ----  ------------------------------------
  132  CYL
  108  KT
  172  TPCR  Track Pitch Correction params
    8  MFGF  Mfg Flags for Servo Manager
2,244  PES   Position Error Signal ?
  116  RWGP  Read/Write Gap table - microjogs
1,012  RW32
  344  RWTC
5,388  SWTG
   24  WIND
7,188  RRO2  Repeatable Runout ?
  456  ACDC   empty
  680  MACT  Microactuator normalization parameters
  260  MAFF  Microactuator feedforward ?
4,928  RRDC   empty
2,056  FLTS   empty

[jono-ats]

I would imagine there are some platter adaptives that have to agree with the ROM. For example, in older HGST products, the NV-RAM (within ROM) points to a boot location on the media.

The only adaptive section that looks like it might point to SA cylinders (or zones ?) is the "CYL" component in the FSVS module. Otherwise there is nothing which looks like the old style "NVRAM". Of course ICBW.

Adaptive sections in FSVS module in HUH728080ALE604 ROM

Code:

size   ID    Description
-----  ----  ------------------------------------
  132  CYL
  108  KT
  172  TPCR  Track Pitch Correction params
    8  MFGF  Mfg Flags for Servo Manager
2,244  PES   Position Error Signal ?
  116  RWGP  Read/Write Gap table - microjogs
1,012  RW32
  344  RWTC
5,388  SWTG
   24  WIND
7,188  RRO2  Repeatable Runout ?
  456  ACDC   empty
  680  MACT  Microactuator normalization parameters
  260  MAFF  Microactuator feedforward ?
4,928  RRDC   empty
2,056  FLTS   empty

Good info. Thanks!