Samsung hdd HD753LJ Dead

[seabird]

This drive died and after following the comments of members here i changed the tvs diodes, this did not work. What i found was even with the 5v tvs removed there was a short acoss the connections, this was not the case for the 12v tvs. My question is does anyone know what would cause this short on the 5v side. Any help would be very much appreciated

[BlackST]

Cause : crap PS. Bad PCB now.

[seabird]

Constructive help is what i need, after reading these forums i know it probably was the PS, although i have not had any trouble with the other drives in the same computer.

[fzabkar]

I'm not a DR guy, but I'm wondering if the preamp inside the HDA gets its positive supply directly from the +5V rail. If so, do you still measure a short with the board removed?

AFAICS, one other possibility is that the +5V rail is used by the motor controller chip to derive the Vcore and Vio supplies for the Marvell MCU. I can see two 4R7 coils, but I can't see any external MOSFETs or flywheel diodes, so I'm guessing that they are internal to the SH6125 chip.

[seabird]

Looks like I`m stuffed might try and get another board. I have got another drive the same but a later one and do not want to destroy both drives by trying the board from that one.

Thanks for your input guys.

[BlackST]

Another boad won't work,usually. And YES, if the preamp is fried you'll have TWO damaged drives.

[seabird]

Thanks BlackST I am listening to you. Bit sick losing a large drive, but there you go.

[guru]

Pre-amp fries when 5v or 12v hits GND

You will have two voltage inputs 5v VCC 8v VDD on most newish TI pre-amps

[BlackST]

Pre-amp fries when 5v or 12v hits GND
You will have two voltage inputs 5v VCC 8v VDD on most newish TI pre-amps

In fact :

I'm not a DR guy, but I'm wondering if the preamp inside the HDA gets its positive supply directly from the +5V rail. If so, do you still measure a short with the board removed?

AFAICS, one other possibility is that the +5V rail is used by the motor controller chip to derive the Vcore and Vio supplies for the Marvell MCU. I can see two 4R7 coils, but I can't see any external MOSFETs or flywheel diodes, so I'm guessing that they are internal to the SH6125 chip.

Most of all, with so many 10$ multimeter with internal 9V battery , in continuity mode or Ohm x 1, it's sure death for preamps AND read channel if you touch the right (or wrong) pins. Bingo !

[fzabkar]

Most of all, with so many 10$ multimeter with internal 9V battery , in continuity mode or Ohm x 1, it's sure death for preamps AND read channel if you touch the right (or wrong) pins. Bingo !

There is no difference between a $10 digital multimeter and a $200 Fluke, in this respect.

All the sensitive inputs would be at the signal end, inside the HDA. I understand that GMR heads are susceptible to damage by ESD, as would sensitive CMOS inputs, but if you seriously believe that a DMM can damage such chips, then you need to learn how a DMM works.

Hint: Use your $10K test equipment, or any $10 meter, to measure the OC voltage and short circuit current of *any* DMM.

BTW, Flukes use 9V batteries, too. Which DMM do you use? ;-)

[BlackST]

I use Agilent. Add one zero to the Fluke price and it's not exactly portable, it's for use on a bench. But doesn't kill anything.
And if you have some scrap drive but with still working preamps, do some tinkering with common multimeters... I have already seen what happens in certain conditions. There should be a reason to say it. Anyway, it's not my problem.

[fzabkar]

I use Agilent. Add one zero to the Fluke price and it's not exactly portable, it's for use on a bench. But doesn't kill anything.
And if you have some scrap drive but with still working preamps, do some tinkering with common multimeters... I have already seen what happens in certain conditions. There should be a reason to say it. Anyway, it's not my problem.

The fact that a Volkswagen is not a Porsche doesn't make it a bad car. They both bring the groceries home.

As for test equipment, I doubt that a $10 DMM is any less capable than your $2K Agilent for the job being discussed. I can see why some people wouldn't trust tools that don't cost enough, but then such people aren't pros. In any case, your benchtop DMM would be powered from 230VAC. Are you sure that it's safe? ;-)

I suspect that your anecdote may just be DR folklore that you've come to accept without question, just like a lot of the Internet "information" that you are fond of ridiculing, sometimes justifiably.

Seriously, I'm curious as to the specs of this meter. What is its OC voltage and short circuit current on the lowest resistance ranges, including the diode test? It should take you two minutes to find out.

[BlackST]

That's what professional instruments are built for. I ask my colleague who has another identical with identical calibration to measure and I'll let you know.
In any case I prefer to put my ass on the Porsche rather than on a VW

P.S. before you say something : even Ferrari is controlled by FIAT, but there's a "little" difference.

[dobrevjetser]

Measuring current is different for these multimeters.
My fluke uses 0,1 mA for diode measurement, whereas a cheap multimeter would use 1 mA.
Enough difference to kill or kill not a sensitive preamp.

Dobre

[BlackST]

In fact. I want to measure with another identical instrument, calibrated. At these level, even the resistance of test leads is important. And if it costs >1700 EUR there should be a reason (beside the ISO calibration that costs a lot by itself).
I'm planning to get a even more sensible instrument for very low ohm measurements BUT with same specs.

P.S. Dobre, what Fluke do you have ? Series 89/189 or bench ?

[fzabkar]

Measuring current is different for these multimeters.
My fluke uses 0,1 mA for diode measurement, whereas a cheap multimeter would use 1 mA.
Enough difference to kill or kill not a sensitive preamp.

Dobre

Both my meters (AU$70 and AU$100) use 1mA. I've buzzed hundreds of ASICs in all manner of equipment without any damage. I don't see why HD preamps are so special. AFAICS, even the sensitive (?) front end needs to be able to handle the 6mA (?) bias current for the GMR heads. Of course, that would be with power applied. Not quite the same thing ...

I also have a Fluke/Philips PM97 Scopemeter which cost me AU$2750 about 20 years ago. It's been a useless paperweight for most of its life. One day I'll either try to fix it again, or step on it. :-x

In any case, the premise that a DMM powered by a 9V battery is inherently dangerous is laughable. That's what aroused my initial skepticism.

BTW, I do know of instances where a DMM can damage a device, although thankfully not from personal experience. EGO sensors are one example.

[fzabkar]

In fact. I want to measure with another identical instrument, calibrated. At these level, even the resistance of test leads is important. And if it costs >1700 EUR there should be a reason (beside the ISO calibration that costs a lot by itself).
I'm planning to get a even more sensible instrument for very low ohm measurements BUT with same specs.

I'm just asking for the OC voltage and SC current. No need for any calibration -- it just needs two minutes of your time. Furthermore, there is no need to measure with another $10K DMM. A result that is within 50% is good enough. A cheap meter will get you within 2%.

If you need help with the measurements, I'll upload a photo.

[BlackST]

If you need help with the measurements, I'll upload a photo.

Noooo... don't worry... stallions know how to make yearlings.

(I just wanted a precise measure for you, I don't trust at all cheap multimeters. Not today, please, I'm "somehow busy" )

[fzabkar]

If you need help with the measurements, I'll upload a photo.

Noooo... don't worry... stallions know how to make yearlings.

Yes, but they only need two minutes to do it.

[fzabkar]

Well, I think a week is long enough time to wait for a stallion to perform.

The photos are here:
http://www.users.on.net/~fzabkar/DMM_test/

If you need any help understanding the procedure, feel free to ask.

BTW, I paid AU$70 for the QM1320 meter about 10 years ago. The same model is still available, for AU$30 (US$28):
http://www.jaycar.com.au/productView.as ... rm=KEYWORD

Attachments