Jump to content

WD - PCB replacement + pictures


Recommended Posts

Since my crystal ball is again in the shop for tuning :(, I have to use I-CHING, that tell me:

I-Ching Hexagram 31

The Judgement

Influence. Success. Perseverance furthers. To take a maiden to wife brings good fortune.

The weak element is above, the strong below; hence their powers attract each other, in your case the weak element is the PCB and the strong one is the HD, the general meaning is that you are doomed to success, but the road to reach it will not be easy.

:P

kaclaz

:P:lol:

I will try the replacment and return with updates :hello:

Have A Good weekend everybody!

Edited by CrazyDoctor
Link to comment
Share on other sites


The design of that power area of hard drives is fairly classic. You usually have 2 TVS diodes, one for the 5v and one for the 12v. They protect against reversed polarity and voltage spikes. So if you don't plug it backwards and that your PSU isn't faulty, the drive will happily work without them (if it's for data recovery only, I definitely wouldn't bother). They're fairly big as well, it shouldn't be hard to remove, even for someone inexperienced. What most likely happened (and that people never want to admit), is you plugged the drive backwards, which will do exactly what we see here (admittedly, it's harder with SATA connectors but lots of times they're on adapters)

As for the resistor, it is indeed a 0 ohm (in the last picture at least), so you could just short the two pads (and nothing else around it, obviously) with solder instead. That is assuming it was plugged backwards, and that only the diodes are bad. If something else is short elsewhere on the PCB, it might not be a good idea, so I'd check the resistance on the 5 and 12v buses (with any old digital multimeter) before I power it on first (if it's near 0 ohm, you DON'T want to plug that in your computer).

Feels like I'm repeating myself ;)

Link to comment
Share on other sites

Great knowledge ! :thumbup

How can I know what is gone on a pcb besides when it's so clear like in the picture?

I mean, is there any technique or a multimeter which can help me to diagnose?

Best Regards,

Link to comment
Share on other sites

You usually have to disconnect one (or more) solder points to use a multimeter for any given component to get correct readings. Even so, some components can't be measured with such a simple device (e.g. transistor); you would possibly need an O-scope and even that may not do. I was in the USN and to repair the Fire Control Console sometimes it required simply swapping complete PCB's to isolate the failures.

I think the "0-ohm" may be slightly incorrect. That would be a complete "short", so it must be somewhere above 0-ohm and below 1-ohm (a fraction). Resistors have a color-striping-combination that is used to cross-reference to a chart for exact values. Many/most other components have a "marking" on them corresponding to the type of component to get the "value".

I still recommend taking the smarter way of replacing the whole PCB, as it was stated above that other components could be bad and IMHO it would become extremely difficult (see first paragraph) to check each and every component.

Again, trust me... I know!

Link to comment
Share on other sites

I think the "0-ohm" may be slightly incorrect. That would be a complete "short", so it must be somewhere above 0-ohm and below 1-ohm (a fraction). Resistors have a color-striping-combination that is used to cross-reference to a chart for exact values. Many/most other components have a "marking" on them corresponding to the type of component to get the "value".

Actually 0 (zero) Ohm resistors are quite common on modern PCB's :whistle::

http://en.wikipedia.org/wiki/Zero-ohm_resistor

I still recommend taking the smarter way of replacing the whole PCB, as it was stated above that other components could be bad and IMHO it would become extremely difficult (see first paragraph) to check each and every component.

Again, trust me... I know!

Never trust the teller. Trust the tale.

:angel

jaclaz

Link to comment
Share on other sites

The resistance is only approximately zero; only a maximum (typically 10–50 mΩ) is specified. Thus, a fractional tolerance (as a percentage of the zero-ohm ideal value) would be infinite and is not specified.
Jeez, dude! Even a WIRE has resistance! Ask any Sound Engineer (of which I am).

Wire Resistance and here.

Edited by submix8c
Link to comment
Share on other sites

How can I know what is gone on a pcb besides when it's so clear like in the picture?

It totally depends. But for anything more advanced than that, you'd need more "specialized" equipment and a fair amount of knowledge. And in a lot of cases, there is just nothing you can do, short of swapping the proper PCB with the right firmware (might require replacing chips) & everything (might even require a clean room). However, in a large number of cases you will see fairly typical failures that are really easy to recognize. Just by knowing the basic layout of that PCB area/placement of the diodes vs power connector/sound knowledge of quintessential electronics designs, you can already tell from the beginning it's a TVS diode, and that D3 is on the 5v rail (if my eyesight doesn't fail me, it's a General Semi/Vishay SMBJ5.0CA), and that when you apply 12v to it, it will do its job (conduct so the voltage doesn't damage anything), and then short (being a silicon device) when it fails, ending up with the result we see here. Applying 5v on the 12v rail won't make the other diode clamp, and D4 seems to be intact as well. It's one of the easiest problems to spot and understand on a hard drive.

I mean, is there any technique or a multimeter which can help me to diagnose?

Well, it's pretty easy going across the different supply rails (3.3v, 5v, 12v) and ground to see if you have a short (or even across the TVS diodes directly; diode check mode works well for this too), but what you can do with a multimeter alone is fairly limited. It's a very good, essential yet basic tool (although you still have to know what you're doing). And it doesn't have to be crazy expensive, I've seen $400+ multimeters that sucked in many ways when compared with models 1/4 of that price. Also, fancy and expensive tools are only useful in the hands of those that know how to use them e.g. a $25000 LeCroy oscilloscope wouldn't have found the problem for you (it's merely a really expensive paperweight in the hands of a n00b)

You usually have to disconnect one (or more) solder points to use a multimeter for any given component to get correct readings

That totally depends what you're trying to accomplish. If you just want to see if various protection circuitry on the power supply side have failed (and similar simple things), you rarely have to.

Even so, some components can't be measured with such a simple device (e.g. transistor)

Actually, a large numbers of multimeters have had that function for a long time ;) Not that it is very practical in most cases admittedly (having to de-solder, SMD parts not having long leads, etc)

you would possibly need an O-scope

I don't see how that would even help here (or with most hard drive PCB repairs for that matter)

I think the "0-ohm" may be slightly incorrect. That would be a complete "short", so it must be somewhere above 0-ohm and below 1-ohm (a fraction)

Of course it's not a perfect 0 ohm. Even a piece of copper/silver that length near absolute 0 degree doesn't have a perfect 0 ohm resistance. It's very close to it however (usually about 0.01 ohm -- something most DMM's can't measure anywhere near accurately).

Resistors have a color-striping-combination that is used to cross-reference to a chart for exact values. Many/most other components have a "marking" on them corresponding to the type of component to get the "value".

I take it you've never worked with SMD parts. SMD resistors don't have color coding. They do have markings "just like other components". A single zero like here, or 3 of them (I see those a whole lot more of those around here) does mean zero ohm.

I still recommend taking the smarter way of replacing the whole PCB, as it was stated above that other components could be bad and IMHO it would become extremely difficult (see first paragraph) to check each and every component.

Again, if it was merely plugged backwards and that the protection did its job like it normally would (and by judging the look of the TVS diodes it seems to have; I'd bet good money that D3 is shorted), then he should be able to get it to work without resorting to that. The TVS diodes are merely protections (read: optional). If after removing them you don't have a crazy short on the 5v or 12v rails anymore, then most likely it'll work just fine (there's always a remote chance damage was done to other components in the picosecond or so before the TVS clamped). It's definitely good enough for data recovery (I personally wouldn't want to keep using that drive long-term, even with a new PCB). Incurring cost (and waiting for parts) to replace the entire controller vs "repairing" it in about 5 seconds for free, just to recover data doesn't necessarily sound smarter to me. Then again, if it's still shorted after doing that, I would get another controller, for sure. Mind you, even that may not work then as one of the sensitive things that might fry if the TVS didn't clamp fast enough is the preamp located on the head stack (good luck changing that yourself!)

Link to comment
Share on other sites

Jeez, dude! Even a WIRE has resistance! Ask any Sound Engineer (of which I am).

Really? I didn't know that.

I would have thought that asking an Electric Engineer would have been more appropriate. :unsure:

However you made a statement:

I think the "0-ohm" may be slightly incorrect. That would be a complete "short", so it must be somewhere above 0-ohm and below 1-ohm (a fraction). Resistors have a color-striping-combination that is used to cross-reference to a chart for exact values. Many/most other components have a "marking" on them corresponding to the type of component to get the "value".

Unless you have some superconducting material, since even a wire has a resistance, the "complete short" you were talking about is a logical impossibility. :whistle:

So, once we have ruled out the existance of a "complete short", there are two COMMON ways to do a "good enough short":

  1. using a short piece of wire or "jumper"
  2. using a 0 (zero) Ohm resistor

Method #2 is COMMONLY used on modern PCB's, for the reasons explained in the given link:

A zero-ohm link or zero-ohm resistor is a wire link used to connect traces on a printed circuit board that is packaged in the same format as a resistor. This format allows it to be placed on the circuit board using same automated equipment used to place other resistors instead of requiring a separate machine to install a jumper or other wire.

BTW - remember that the metrics you used in the past:

http://www.msfn.org/board/index.php?showtopic=142620&st=38

may also apply here. ;)

:P

jaclaz

Link to comment
Share on other sites

Great things were been said here! :thumbup

I am a first year student of electricity and electronics engineering.

It's interesting when we will get to the part of all what you said above here.. :rolleyes:

What tools are you recommend to me to get, for data recovery?

Edited by CrazyDoctor
Link to comment
Share on other sites

Since you're a first year electronics student, I suggest you just stick to

a multimeter and an adjustable power supply.

Maybe, only maybe, you are even able to get yourself a nice cheap second hand

oscilloscoop, but it might be better to ask Santa for that at a later time,

perhaps when going to second grade. ;)

At school you will learn how to use these two pieces of equipment in due time.

This doesn't help you starting a data-recovery lab of your own, but it does

give you a good starting point of becoming an electronics engineer.

Second tip is to just examine every PCB you own intensively, trying to figure

out what does what (it would help a lot if you'd have the schematics, though).

Like CoffeeFriend says: equipment isn't everything; you can go a long way by only

using a multimeter (once you know how to use it), but don't expect miracles: it's

the man behind the equipment that does the determination of the problem.

After that comes the part of fixing the problem, which needs some skills that can

only be done by practicing often.

And don't forget: a lot of failures lead to success :angel

Greetz,

Peter.

Link to comment
Share on other sites

I am a first year student of electricity and electronics engineering.

It's interesting when we will get to the part of all what you said above here.. :rolleyes:

Everything I said is really basic stuff. You have a long journey ahead of you :) Electronics today aren't anything like the "old days" (based on 555's, 741's, 74LS series, etc). It's one of the most rapidly changing fields (that, and programming -- it kind of goes together). There is a whole new world beyond simple electronics... From microcontrollers to DSPs, to CPLDs and FPGAs and so on (and everything else that goes with it: C/C++ code, Verilog/VHDL, etc) I would normally recommend getting "The Art of Electronics" by Paul Horowitz but it's over a decade old, and the 3rd ed should be out this year.

What tools are you recommend to me to get, for data recovery?

I am not a data recovery specialist (I have very little interest for that), I just happen to work in the embedded electronics field. So I can't be of too much help here. I've had good luck with R-Studio personally. I would be better at answering electronics-related questions ;)

Since you're a first year electronics student, I suggest you just stick to a multimeter and an adjustable power supply.

That covers the basics :) There is no way around having a decent multimeter. As for PSU, lots of times we can get by with a spare ATX PSU rigged with a switch and a dummy load, but sometimes you just need adjustable (although a classic LM317T will do in a pinch) and dual polarity...

Maybe, only maybe, you are even able to get yourself a nice cheap second hand oscilloscoop, but it might be better to ask Santa for that at a later time, perhaps when going to second grade. ;)

I wouldn't personally bother as he will likely have access to one in labs anyway. And something nice isn't always cheap. If I was going to buy one now (on a budget), I'd likely go for a Rigol DS1052E (not as well known a brand as Tektronix, HP/Agilent or LeCroy for sure; however they are the ones who manufacture Agilent scopes!) for around $400. Then again, there are so many other things one could use... A decent soldering station is a must -- doesn't have to be fancy or expensive, even a classic Hakko 936 will do just fine (dirt cheap, a large number of tips are inexpensive and easy to find, etc). I could make a fairly long list :)

Anyway. Good luck to you.

Link to comment
Share on other sites

Electronics today aren't anything like the "old days" (based on 555's, 741's, 74LS series, etc).

You have a most queer idea of what "old" means. :w00t:

"My" old means:

http://en.wikipedia.org/wiki/Vacuum_tube

AL3, AL4, etc.:

http://www.qsl.net/dl7avf/roehren/index.html

and, at the most:

http://en.wikipedia.org/wiki/Transistor

BC107, BC108 2N3055.....

:P

jaclaz

Link to comment
Share on other sites

You have a most queer idea of what "old" means. :w00t:

Well, it is fairly old. The 555 is from 1970, the 741 is from the 60's, and the 74 series would be from the 60's as well. That's universally older than the original PC XT (which I would also call old, although I've had FAR simpler computers too), and older than a sizable amount (majority?) of our forum members.

As for tubes, I've mostly used them in RF applications. I'm not as familiar with them mostly because I'm not one hundred years old yet ;) And to have used them outside of audio applications such as digital logic, you pretty much have to go back to the ENIAC era in the late 1940's. I would perhaps have included a couple on the list if this post was read by folks in retirement homes :P *ducks*

and, at the most:

http://en.wikipedia.org/wiki/Transistor

BC107, BC108 2N3055.....

BC* BJTs never got popular in North America. That was mainly a source of frustration and hours wasted searching into a ECG or NTE xref book (no fancy internet back then obviously). Here you'd be talking about 2N3904's & 2N3906's, and to a lesser extent the 2N2222. I left those off this list precisely for this reason: I knew people outside of NA might be more familiar with other stuff. Although the 2N3055 is sure one of the classics around here too (when more power was involved, that was pretty much the first thing you'd think about). Also, I figured a lot more people would likely know about classics such as the 555 (I think even my neighbour's dog knows about this) than specific transistors.

I've worked with much older stuff like core memory, or programs that would load from tapes (which were almost as big as a VHS tape) at 50Hz (which includes parity and start/stop bits), systems built entirely in wire-wrap and such. That's not old, that's older than dirt :lol:

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...