I had two drives over voltaged and I managed to fix one of them following fzabkar tutorial.
On my second drive which is a 3TB WD30EZRX I have tested D3, D4, R64 and R67
testing on 200 ohm:
D3 shows 176 both ways
D4 shows 1 both ways
and both resistors display 0.
I’m unsure which is the faulty component as the drive won’t even spin up and won’t cause the PSU shutdown.
scorp200 wrote:
On my second drive which is a 3TB WD30EZRX I have tested D3, D4, R64 and R67
testing on 200 ohm:
D3 shows 176 both ways
D4 shows 1 both ways
and both resistors display 0.
***********
Are you testing them in-circuit or removed from the circuit? It makes a big difference. Even lifting one end will give a correct measurement. If there is a printed circuit and you can’t get at the devices because they are surface mount, you can carefully cut the printed circuit trace at one end of the device.
I presume the D3, D4 are diodes. Out of circuit they should measure about 0.7 one way and infinite the other way.
If you are measuring in-circuit, all bets are off. There are meters that will measure in-circuit but the average DVM wont.
It’s highly inlikely that two resitors will test 0 ohms unless they show sign of over-heating. You probably have something in parallel with them that has a very low resistance. That’s because you are measuring in-circuit.
How did you get an over-voltage situation in a circuit with regulators? Power line spikes are handled well with computerized equipment and I can’t see how you got too high a voltage on a hard drive.
i took off the PCB from the drive completely then tested it, and I had a modular PSU and it seems like i plugged in the power suply for the drives into wrong socket. and yes the D3 and D4 are diodes and the resistors in quesion are 0 ohm resistors.
edit:
if i run a 200 ohm test both resistors show 0.5
and for 2k test D4 actually shows 0.787 one way and “1.” other way
if i run a 200 ohm test both resistors show 0.5
and for 2k test D4 actually shows 0.787 one way and “1.” other way
0.5 ohms sounds ok for a resistor but normally such a low value is used with power devices and the resistors are relatively large in physical size. With a very low resistance like that, any amount of voltage across the resistor will cause high currents, so the power rating of the resistor has to be high.
Also, you could be dealing with a fuse, or fusible resistor.
With digital voltmeters (DVMs) you can get away with a 200 ohm range but you should really get onto as low a range as possible if your resistors are that small.
If you have fried them due to over-voltage, there should be signs of burning. If you get your nose close to them you’ll smell it, or the PCB will have signs of burning on the traces. The traces may even melt. Get a high powered magnifying glass and have a close look. However, a high powere transient voltage will take out devices without showing damage. It depends how long you had the over-voltage applied.
With a diode measurements of 0.787 one way and 1 the other way, I am sure you have a device in parallel, like another diode. When you reverse bias a diode junction it should read almost infinite resistance.
Removing the PCB is not enough. You have to interrupt the diode circuit at one end.
I have spent hours and hours trying to repair hard drives and the returns are disappointing. I have opened them up, removed the heads and cleaned them but the tolerances are so incredibly small these days that getting them working again without the proper equipment is virtually impossible.
I have worked in electronics for decades and repairing a hard drive is far more complex than checking a few resistors and diodes. I would never try to discourage you from experimenting but I can tell you from experience that an over-voltage condition has likely taken out junctions inside chips and there’s no way to repair those, even if you can find them. Junctions in transistors are far more fragile than resistors and they’ll go just by sneezing on the chip.
The problem with chips is finding the right one, then there is the problem of changing a surface mount chip. The diodes and resistors you are testing are likely only an 1/8" long, if that.
there is no burn marks at all on the PCB, it is clean. and how do I interupt a current to a diode?
i removed D3 and fused its resistor and the only reaction i get it disconnects my second drive for a second,
but this one will not spin up.
scorp200 wrote:
there is no burn marks at all on the PCB, it is clean. and how do I interupt a current to a diode?
i removed D3 and fused its resistor and the only reaction i get it disconnects my second drive for a second,
but this one will not spin up.
you test diodes and transistors with an ohmmeter. Assuming the red lead goes to the positive of the meter’s battery, if you put the red lead on the diode’s anode and the black lead on the cathode (normally marked with a bar), you will forward bias the diode and the meter will read about 0.7 ohms. If you reverse the leads, you reverse bias the diode and it should read infinite, or very high.
You need to disconnect one end of the diode from the circuit, otherwise current from the ohmmeter can go through parallel components.
The ohmmeter has an internal battery that pushes a current through the load it is measuring and the meter. Voltmeters and ammeters don’t use batteries, they draw current from the source they are measuring to operate their meter or digital readout. The ammeter uses a low resistance shunt in parallel with the meter and a voltmeter uses a resistor in series with the meter.
Be sure to use a modern ohmmeter, like on a DVM, if you use it to test transistors. The older meter provide too much current and it can blow out the junction.
I would not short out components unless you have a circuit diagram and know exactly what is going on. Disk drives have complex circuits to control motor speed and positioning. They also have a built in processor and BIOS. If you short components there’s no telling what effect you will have on the circuitry.
Also, you take the chance of over-loading the power supply and blowing it. The power supply has a certain amount of protection but it is also a toroidal device that is sensitive to transients. If you start shorting components you could do serious damage right down the line and possibly damage your motherboard as well.
I don’t know what to do now, the PCB it self doesnt show any damaged parts and removed the D3 diode and shorting the resistor next to it give some feedback to my computer by disconnecting other drives for split second… but the drive it self wont spin up.
could it be a problem with the SATA/power connection it self?
If you can upload a detailed photo or scan of your board, I can identify the voltage test points for you.
Otherwise, if you need to replace the PCB, then you will need to transfer the flash memory chip at U12 (or its contents) from patient to donor. This chip stores unique, drive specific “adaptive” information.
The following PCB suppliers include a free transfer service:
this is before i removed D3 and shorted R67
Not enough detail. Sorry, I can’t help you.
you need close ups and back of the board?
fzabkar wrote:
if you need to replace the PCB, then you will need to transfer the flash memory chip at U12 (or its contents) from patient to donor.
It’s not that simple. That chip contains timing data for a specific drive and the data is written after the drive is assembled by a very expensive alignment machine. If you swap the chip between drives there’s no reason why the drive should work properly.
There is Russian software and hardware, I think it is the PC3000, that can help align drives that are not too far out of alignment but last time I checked they wanted more than $1000 to buy the software/hardware package in the US. The package is a lot cheaper in Russia.
If you have information on how to successfully transfer the EEPROM chip between drives I’d be interested in hearing it.
@gordo999, I’m not suggesting that you swap U12 between DRIVES. Why would anyone do that? Instead I’m saying that you need to transfer U12 from patient PCB to donor PCB. In fact that’s standard practice.
As for how you would transfer the “ROM”, you would use a desoldering station or hot air rework station. Alternatively, you could use an SOIC chip clip and a device programmer. Lots of info at HDD Guru and The HDD Oracle …
BTW PC3K doesn’t “align” drives. In fact I have no idea what you mean by that.
As for the cost, Ace Labs want a lot more than $1K. It’s more like $10K. That said, there are much cheaper tools, eg SalvationData’s HD Doctor or DFL-WDII. I think they’re around $600. If you go to HDD Oracle, there will be other tools, some of them “free”.
To scorp200, I need a high resolution photo of the component side. The reverse side is uninteresting for our purposes.
Ideally I’m hoping for this kind of shot:
http://malthus.zapto.org/download/file.php?id=276&mode=view
Or these:
http://malthus.zapto.org/viewforum.php
You might find these tutorials useful:
Tutorial - Linear and Switchmode Regulators used in HDDs:
http://malthus.zapto.org/viewtopic.php?f=59&t=231
Analysis of the “ROM” on a Western Digital ROYL HDD:
http://malthus.zapto.org/viewtopic.php?f=59&t=225&p=548
SOIC 8 Clip (and in-circuit programming):
http://malthus.zapto.org/viewtopic.php?f=110&t=85&p=1233
fzabkar wrote:
BTW PC3K doesn’t “align” drives. In fact I have no idea what you mean by that.
*************> Inside the EEPROM there is data put there by a specialized machine that corrects the motor speed and the seek stepper motor. The machine reads the alignment and provides correction parameters to adjust the head stepper motor and drive motor for an individual machine. Those paramters are recorded in the EEPROM. The parameters are read by the controller logic and used to position the head dynamically.> When you are dealing with very fine tolerances in seeking and reading data at variable speeds (the disk angular velocity is different at different spots on the disk and the data density varies as well according to position) you need a dynamic arrangement that will adjust the head seeking functionality. That fine-tuning info is stored in the EEPROM. If you swap PCBs, EEPROMS, or disks, the fine-tuning will be different.> The PC3K can reset some parameters that affect how the drive performs. I am not claiming it can align heads physically but it can alter the fine-tuning parameters that position the head dynamically. The only way to align the heads physically is to adjust them physically inside the drive, and good luck doing that.> At the factory, once the basic alignment is adjusted, the drive is sealed and the mechanism is accessed via a special machine that attaches physically through a port hole in the drive. After final alignment, the port is sealed. The data acquired by that special machine related to alignment is stored in the EEPROM.> Sometimes a drive becomes useless because data gets written to the area on disk reserved for storing such data. That can happen during a power fail if the head is writing and suddenly the power to the stepper motor is cut, or the power to the drive motor.> The drive logic is lost during a sudden power transient and the heads can end up anyhwere while still writing. If the onboard disk data in the special maintenance area is over-written, or the heads go slightly out of alignment, PC3K can be used to correct ‘SOME’ alignment problems related to damaged data.
There is nothing in the EEPROM to “correct the motor speed”. The RPM is sensed and controlled by the motor controller. See the SMOOTH L7250 datasheet. The EEPROM would in fact contain the initial register settings for the motor controller, and one of these registers would have to set the desired RPM, but that’s not an “adaptive” setting. RPM is a fixed parameter that is the same for all drives of that particular model.
There is no “head stepper motor”. Stepper motor positioning systems disappeared in the early 1990s and were replaced by voice coils (VCM). Once again, see the SMOOTH L7250 datasheet. In fact WD were one of the last manufacturers to transition from steppers to voice coils.
The EEPROM stores “adaptive” parameters. If you replace the original PCB with one from a different drive, your drive will inherit the other drive’s adaptives. That’s why you need to transfer the adaptives from your drive’s PCB to the replacement PCB. You do that by transferring the EEPROM, or its contents.
The servo track information is laid down by a Servo Track Writer via a push-pin. Subsequently the drive is calibrated to take account of the differences in the characteristics and performance of each head. Some heads will have better frequency characteristics, so their respective tracks will be recorded with greater bits per inch. The gain of each head and the head bias current will also be tuned for the minimum error rates, and the microjogs will be adjusted to account for the different physical spacing between each head’s read and write elements.
WD has factory software called Trex. It performs ARCO (Advanced Read Channel Optimisation), ASCO (Advanced Servo Channel Optimisation), and PERCO (PRML Enhanced Read Channel Optimisation).
The following article was written by a Seagate employee involved in data recovery. The article should also apply to WD’s drives. It explains the reason for the adaptives.
HDD from inside: Tracks and Zones:
http://hddscan.com/doc/HDD_Tracks_and_Zones.html
The following patent document has a good explanation of microjogs:
http://patentimages.storage.googleapis.com/pdfs/US5978168.pdf
| Forums | News and Announcements | Register | Help | Forum Guidelines |
Copyright © 2021 Western Digital Technologies, Inc. All rights
reserved.
|
Trademarks
|
Privacy
|
Community Terms of Service
|
Site Terms of Use
|
Copyright
|
Contact WD
|