Hitachi HDS723020BLA642 always different sounds - diagnosis?

[michael chiklis]

In any case, do you really think you have a PCB fault?

Yes, because as i wrote into 1st post, the drive bricked after power surge.
But i don't know if there is also internal faulty due to power surge.

If i will find a donor pcb i will do NVRAM transfer with my hot air soldering station, eeprrom programmer is too risky!
However, tomorrow i will try if i can find 18V with pcb connected to the drive.
Thank you.

[michael chiklis]

Hello,
i didn't find any +18V with pcb connected to the drive.
Also, as you can see in the pic i posted before, on 47 ohm resistor i have 5V on both side.
Resistor R470: 47x 10^0 = 47 ohm

This means it is shorted.
So i checked the resistor with ohmeter, value is 0 (shorted)!
Aunless the resistor is not connected in parallel to the coil, it should not be 0 ohm.

[michael chiklis]

Also, i found 1 opened fuse (if it's really a fuse)?

[fzabkar]

The "R' in "R470" is a decimal point. So the resistance is 0.47 ohm.

Those other two components should also measure 0 ohms. That said, if the 5V "fuse" is open, then why is the 5V TVS diode not shorted? Is it possible that the surge struck the 5V supply, in which case is it possible that it killed the preamp?

[michael chiklis]

That said, if the 5V "fuse" is open, then why is the 5V TVS diode not shorted? Is it possible that the surge struck the 5V supply, in which case is it possible that it killed the preamp?

That's exactly what i was thinking...
5V diode is not shorted, it should short to protect other pcb components and pre-amp.
On seagate drives i see TVS diodes do a good job, but on WD (green models) i see very often TVS protection failing.
This is the first 3.5" Hitachi on which i'm working, and TVS protection seems to be as WD, SUCKS.

i fixed few days ago a 2.5" Hitachi, it had one open fuse, i substituted it after searching for shorted components as TVS diods, capacitors etc but i didn't find any shorts.
That drive is working since 1 week.

I would love to ask to WD and Hitachi engineers why they do not take example from Seagate (at least as regards the TVS protection)


I was thinking also, why does the drive spin if that fuse is open?

[fzabkar]

I guess I've been in denial for a long time, but ISTM that, as strange as it may sound, Hitachi's (and WD's ?) design philosophy seems to be to protect the external PSU from the consequences of a shorted TVS diode rather than to protect the PCB from an overvoltage. In other words, it appears that Hitachi's circuit may look something like this:

Code:

+5V  O----+---------O +5V unfused
          |
         fuse
          |
          X <---- does this junction connect to anything?
          |
      TVS diode
          |
GND  O----+---------O Ground

If I'm right, then the first surge opens the fuse, and the second surge then kills the unprotected drive, or the first surge can do all the damage on its own if it persists for long enough.

FYI, here is a recent thread involving another Hitachi model:

viewtopic.php?t=26543&p=181611#p181611

The TVS diodes appear to be fused on the grounded side. In that case the drive didn't spin up, but there were still voltages present, even though the motor controller had pinhole damage. Unfortunately the OP in that thread declined to provide additional measurements which may have illuminated the problem.

BTW, I would check whether junction "X" connects to any pin on the motor controller.

[fzabkar]

i didn't find any +18V with pcb connected to the drive.

FWIW, the SMOOTH L7250 datasheet states that ...

"Boosted gate drive for the high side drivers [MOSFETS in the VCM Power Driver H-Bridge] is provided by the charge pump circuitry, with the boosted voltage at the VCP pin."

The typical boost voltage is 18.5V.

I don't know how this translates to other motor controllers, but ISTM that you should be measuring a boost voltage that is significantly higher than the +12V supply rail. That is, if I've interpreted this area of the circuit correctly. BTW, my first guess at the circuit layout was wrong. You should be looking for 18V at the diode, not those 4 holes.

[michael chiklis]

as strange as it may sound, Hitachi's (and WD's ?) design philosophy seems to be to protect the external PSU from the consequences of a shorted TVS diode rather than to protect the PCB from an overvoltage.

Well, i think this is stupid decision of Hitachi and WD if they really think to protect external PSU.
All modern computer PSU (switching ATX) have overcurrent and short circuit protection. When protection on PSU gets activated due to hdd pcb shorts, computer doesn't even turn on until you unplug the faulty hard drive pcb and switch off the PSU switcher for few seconds.

I would recommend to Hitachi and WD to think more about how to better protect their own pcb rather than the computer's power supply, assuming that everything is not simply due to their low quality circuit design.

[michael chiklis]

BTW, I would check whether junction "X" connects to any pin on the motor controller.

Nope for the bad fuse (unless between junction "X" and motor controller pins there is some opened component).
Yes for the good fuse

P.S.
Bad fuse is on 5V rail, the good fuse is on 12V rail.

[michael chiklis]

I didn't find any 18V, but i found 24V point.

When i power on the drive i have 12V in that point, then the drive does some clicks (when i hear clicking the tension goes up to 24V for half second).
If drive spins down the tension stay on 12V, if spinning continues it keeps on 24V
I have 24V only in that point and only if the drive keeps spinning.

[michael chiklis]

I changed fuse, but i still have the same issue.
Power surge has killed something else, maybe motor controller chip or maybe head pre-amp (hope not the last one)
I will search for a donator pcb, thanks.

[michael chiklis]

What kind of diode is this?

above it is written:
ST VU
SUA
GP03S

I can't find datasheet for it!!

I dessoldered it from pcb to misure with multimiter.
Diode tester in direct polarization: 0.270 V
Inverted polarization: 0.585 V

20K ohm tester in direct polarization: 2.7 Kohm
2K ohm tester in direct polarization: 1.31 Kohm

20K ohm tester in inverted polarization: 3.98 Kohm
2K ohm tester in inverted polarization: "full scale range"

Drive still spins without the 5V diode and without 5V fuse

[fzabkar]

Diode.JPG

What kind of diode is this?

above it is written:
ST VU
SUA
GP03S

I can't find datasheet for it!! :?

I dessoldered it from pcb to misure with multimiter.
Diode tester in direct polarization say: 0.270 V
Inverted polarization: 0.585V

20K ohm tester in direct polarization says: 2.7 Kohm
2K ohm tester in direct polarization says: 1.31 Kohm

20K ohm tester in inverted polarization says: 3.98 Kohm
2K ohm tester in inverted polarization says: "full scale range"

It's a 5V TVS diode. It is testing as leaky, so it is definitely bad.

"ST" is the manufacturer, ST Microelectronics.

"SUA" is the marking code, probably similar to SMA6J5.0A (6UA marking).

http://www.st.com/web/en/resource/techn ... 152688.pdf

"GP03S" is a date code and manufacturing location.

Drive still spins without the 5V diode and without 5V fuse :shock:

All drives work without their TVS diodes.

[michael chiklis]

Yes, but also fuse has been dessoldered!
Do you think more that the the pre-amp is damaged or the motor controller chip?