WD205AA giving some more trouble

[shaun]

Hi all,

I have a WD205AA here (from a Mac, with an rotten apple logo printed onto the WD label) giving some pain. At first it looked like a dead preamp (no data on the scope), but after a swap it continues clicking. Before I start any further attempts in measuring, I'd like to hear your opinion. I have been given a head stack from a DCM xxxxEQ drive, so it matches mine, BUT it has 4 instead of three heads (0123 instead of -123). I am willing to believe that the stack is from a refurbished drive and actually had a dead No 1 head, which would be the system surface in a -123 config (right?)
Any other issues with those drives which render it useles to try any further?
Is there anything like a terminal (as on Seagate) on WD disks?
And no, I've no PC3000 or something like that on my site.

[harddrivespecialist]

I suggest that you send it to someone else

[shaun]

Yeah, very helpful. You're one of those guys telling his kids to ask someone who can read instead of sending them to school, right?

If you know from experience that 99% of this particular drive model cannot be repaired without tools like PC3000 and therefore unrealistic, you're invited to tell me.
Otherwise just don't write, it simply sucks when some people keep on and on telling others that they are too dumb to learn and only the so-called pros can turn a screw the right way.

[harddrivespecialist]

I did not call you any names!

You should not be guessing with a heads.
You have to match the heads exactly. There can't be "I am willing to believe that the stack is from a refurbished drive and actually had a dead No 1 head"
You have to know for sure which head is working and which is not.

[thatdellguy]

No disrespect but from the way you ask your questions I gather that your new to the data recovery scene. Without special tools you're guessing at the cause of the failure. And you're using donor parts from a incompatible drive. Swapping a four head for a three head drive speaks volumes about your skill level. If the data is important send it to a pro.

[shaun]

Ok, got the point.
I'm experimenting with this one, and I did actually have success in another case with a higher head count on the replacement stack. For that one, I got the data sheet from the preamp and decided that it makes no difference whether head 0 or whatever it was then is present or not, the multiplexing was the same for 3 or 4 heads, the missing head's input was simply unconnected (or connected to ground).
For this experiment, I relied on the 5th and 6th letter of the DCM, which is obviously not enough.
But as one says "learning by doing" and not "learning by letting do", I'd still like to know what problems would occur by swapping a 3 head stack against one with 4 heads, as long as the head selection works the same way and the heads are all in good condition.
I'm an electronics engineer and therefore interested in the electrical differences; for the mechanical and DR related stuff I'd better read what you say and continue learning.

[BlackST]

Did they teach you something about head mapping?

[shaun]

Who is "they"?
As long as I'm dealing with the head stack / preamp combo, I'm completely working in the hardware domain, so only physical mapping would be relevant - from my understanding.
As long as the physical heads of both stacks are connected to the same physical inputs of the preamp multiplexer and the latter is adressed in the same way, I don't see any reason why it should not work.
BUT as it does not work, there are three possibilites left: 1. I'm wrong and the heads are not connected the same way or the preamp IC does have some kind of adaptives (e.g., an EEPROM telling the multiplexer which logical input should select which physical head), 2. the SA on the platter is fubar, or 3. one of the three heads I need is bad on the replacement stack.
I'd agree that I's better have carried out this test on a drive which was working prior to my experiments, but now that I'Ve started thinking about the physical and electrical aspects of heads, I would prefer to learn from it.

[msurgeon]

You're correct that using a 4-head for a 3-head assembly in not necessarily a bad thing. However, if the heads are from a refurb then there is a possibility that one of the heads is bad and is disabled in the head map. You would need to know the head maps on the original drive and donor to determine if the corresponding heads on the donor are operational. Note that heads are numbered from the bottom up- 0 1 2 3. Without PC3000 or similar tool it is not an easy thing to determine the head map.

[shaun]

Thanks for your reply. I've read some GMR preamp data sheets meanwhile. They do have a serial control interface commonly, but no means to store adaptive information internally. The bottom head is missing in my drive-under-test, so I only have physical heads 123.
Would PC3k the only way to tell the mapping for WD drives?

[msurgeon]

The head map is contained in flash ROM on the PCB. Without the PC3000 or similar tool (which can read the ROM in place) you would need to de-solder the flash ROM chip and read the file in a programming device. Deciphering the head map from the file is the tricky thing. I am not aware of any serial diagnostic interface on WD drives; all WD diagnostics on the PC3000 are through the ATA interface.

[shaun]

Ok, I see. But as long as I swap head stacks with the used heads in proper condition, I wouldn't have to worry about head mapping, would I?

[msurgeon]

Correct, assuming there are no other issues involved such as head alignment-which can vary slightly from one set of heads to another- or surface defects. This assumes that you've got a compatible head set to begin with: if they appear to be visually identical and the pre-amp chip matches then it's likely they are. WD has a habit of changing head configurations even within the identical model. Make sure donor is from approx. same date of manufacture as original.

[shaun]

Thanks, I'll give it another try as soon as I find a compatible head stack. This time I'll look explicitly for three heads and non-refurbished HDDs. My assumption that the head stack or the preamp may well be the problem was that I compared the signals on the HDA connector from my disk with those of a working disk with the same PCB. First I assumed that two pins must be the VCM drive, at least one is GND, one is +5V for the preamp from a 7805-type regulator, two or three will be used for the serial control port and two times two will be the differential read and write channels. While I got an approx 0.2Vpp signal on two pins of the good disk, my candidate showed a flat line.

[pepe]

HeadMap is in ROM, so U can read that out and check the head map.

pepe

[shaun]

...which would imply that I knew the position of the head map in ROM, the meaning of the data and an adapter for the flash chip (which would be no problem at all, but is the map stored in the external flash or does the WDC custom chip also contain ROM?)
If you could help me with that, I'd try to read the flash and probably learn something more - which would keep me from swapping different head stacks again