EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: 2X on September 05, 2023, 07:21:34 pm
-
Hello,
I have a Seagate ST3500312CS where is dead because overvoltage of the power supply. When I unscrew the PCB I saw that three capacitors (or maybe ferrite beads) where blown out. Do you have any idea if I have a possibility to fix it. The problem is that I have an industrial custom PC where recognizes only the disk with the certain only firmware (it is crazy but it is true). Does anyone know where the firmware of the disk is saved (I have another Segate disk but with different firmware and I thought If I could replace the firmware chip in the good one). Anyway, I would like to fix it (always there is something new to know in repairs), so there is any schematic of HDDs or what to suggest me to check it will be very helpful?
P.S. I hope I didn't post on the wrong thread.
-
I dont think this is going to work. That chip will also have calbration data for the disks. You cant transfer it to a different drive
-
Are you sure that PC requires certain only firmware? Can it be serial number lock? Back in the day I used hdd serial number lock for some of my software.
Anyway, the firmware may be stored together with head calibration data, so I would say switching chips, even for exactly same model is unlikely to work. If the model differs, it's even less chance. That's my opinion.
If 12V rail is the only one which suffered, the best aproach would be to try repairing. First I would measure the rail for shorts. Maybe the board is still ok. Or maybe you will find that motor driver IC or something else simple is damaged. These should not be hard to replace, and it has nothing to do with calibration (as far as I know).
-
Are you sure that PC requires certain only firmware? Can it be serial number lock? Back in the day I used hdd serial number lock for some of my software.
Anyway, the firmware may be stored together with head calibration data, so I would say switching chips, even for exactly same model is unlikely to work. If the model differs, it's even less chance. That's my opinion.
If 12V rail is the only one which suffered, the best aproach would be to try repairing. First I would measure the rail for shorts. Maybe the board is still ok. Or maybe you will find that motor driver IC or something else simple is damaged. These should not be hard to replace, and it has nothing to do with calibration (as far as I know).
I will try to repair it and not to risk it... by the way which chip contains the calibration data/firmware? Also, these three components that blown out do you believe are capacitors of ferrite beads?
-
Take the original board and remove the TVS diodes. They usually short out protecting the rest of the board.
Removing them will tell if there's permanent damage. The ferrite beads can be shorted out for the test, better put a 1-2Amp fuse or a 0 ohm SMD resistor.
Also, I've repaired several disks by transferring the original eeprom chip to the new donor board.
Make closer pictures of these parts.
-
by the way which chip contains the calibration data/firmware? Also, these three components that blown out do you believe are capacitors of ferrite beads?
I guess firmware is in one of these SO8 packages. SPI flash. Not sure and can't read from picture. The three components should be ferrite beads. It's easy to check with multimeter, if they go in series with supply, they are ferrite beads.
-
I found the one from the SMD code that is maybe Transient Voltage Suppressor. The one soic chip is "25X40CL" and the other one is "NTMD4184PF" I beleive (MOSFET – Power, Single, P-Channel, Schottky Diode, Schottky Barrier Diode). I wll measure all the components that you told me and power supplies and I will tell you the results.
QA NSA5.0AT3G SMA 2 400W Peak Power Zener Transient Voltage Suppressor
http://markingcodes.com/search/c/QA (http://markingcodes.com/search/c/QA)
25X40CL
https://datasheetspdf.com/pdf-file/1315797/Winbond/W25X40CL/1 (https://datasheetspdf.com/pdf-file/1315797/Winbond/W25X40CL/1)
NTMD4184PF
https://www.onsemi.com/pdf/datasheet/ntmd4184pf-d.pdf (https://www.onsemi.com/pdf/datasheet/ntmd4184pf-d.pdf)
-
You missed the other large diode!
What do they measure in diode mode?
-
You missed the other large diode!
What do they measure in diode mode?
You were totally right. I first I measured the "pE49 LEK" and was short... I desoldered and I measured it on the fly again and it was short and also the short from the PCB was gone. Then I measure one other smd component at the side of the sata power connector from another Seagate disk and it was a ferrite bead. I put instead of ferrite bead 1.5Ω resistors and the disk works fine and the machine too. Because I want to repair it as it was, I would like your opinion about what components were the originals.
Resistor: thick film; SMD; 0603; 1.5Ω; 0.1W; ±1%; -55÷155°C
Manufacturer part number: RC0603FR-071R5L
https://www.tme.eu/en/details/rc0603fr-071r5/smd-resistors/yageo/rc0603fr-071r5l/ (https://www.tme.eu/en/details/rc0603fr-071r5/smd-resistors/yageo/rc0603fr-071r5l/)
The good ferrite bead I measured it with keysight U1733C and it shows inductance 1.7uH (I don't find a ferrite bead with inductance characteristic on sites - it is something wrong?) and resistance 1.21ohm. The Seagate disk draw 0.155A @ 12V. Because 0.155A will not flow only through a bead but will divide through three beads. Ibead=0.155A/3=0.052A(ideally that the current is divided equally), so 1.5x0.052A~0.078A=78mA (I found at 100mA).
Part Number: BLM18HD471SN1D
https://gr.mouser.com/ProductDetail/Murata-Electronics/BLM18HD471SN1D?qs=WYZ1yQOz%2FLFGae%2Fi4Y5yJw%3D%3D (https://gr.mouser.com/ProductDetail/Murata-Electronics/BLM18HD471SN1D?qs=WYZ1yQOz%2FLFGae%2Fi4Y5yJw%3D%3D)
The "pE49 LEK" I found on the below site that propably is this part number "NSB12ANT3G"
http://markingcodes.com/search/c/LEK (http://markingcodes.com/search/c/LEK)
https://datasheetspdf.com/pdf-file/691810/ONSemiconductor/NSB12ANT3G/1 (https://datasheetspdf.com/pdf-file/691810/ONSemiconductor/NSB12ANT3G/1)
The below is a good choice for replacement?
Part Number: SZ1SMB12AT3G
https://gr.mouser.com/ProductDetail/Littelfuse/SZ1SMB12AT3G?qs=sK5eIuwOod4%2Fto4W2E8Wrw%3D%3D (https://gr.mouser.com/ProductDetail/Littelfuse/SZ1SMB12AT3G?qs=sK5eIuwOod4%2Fto4W2E8Wrw%3D%3D)
Thanks you and the other guys for the help.
-
The "pE49 LEK" I found on the below site that propably is this part number "NSB12ANT3G"
The code is "LEK", forget the rest.
Quick answer from: https://community.wd.com/t/hdd-tvs-diode-faq/14692
Some common marking codes for 12V diodes are LE, LEM, LEK, BUX, KVP, 13L.
Some common marking codes for 5V diodes are HE, QE, QA, AE, 5L.
Suitable substitutes for the 12V and 5V TVS diodes in 3.5" drives are SMBJ12A and SMAJ5.0A.
The Seagate disk draw 0.155A @ 12V.
Yeah but it will be a lot more at spin-up! But if it's working with 3x 1.5R resistors, I guess it will also work with 3x 1.28R chokes.
I'd try to find others with lower resistance, though.
-
The "pE49 LEK" I found on the below site that propably is this part number "NSB12ANT3G"
The code is "LEK", forget the rest.
Quick answer from: https://community.wd.com/t/hdd-tvs-diode-faq/14692
Some common marking codes for 12V diodes are LE, LEM, LEK, BUX, KVP, 13L.
Some common marking codes for 5V diodes are HE, QE, QA, AE, 5L.
Suitable substitutes for the 12V and 5V TVS diodes in 3.5" drives are SMBJ12A and SMAJ5.0A.
The Seagate disk draw 0.155A @ 12V.
Yeah but it will be a lot more at spin-up! But if it's working with 3x 1.5R resistors, I guess it will also work with 3x 1.28R chokes.
I'd try to find others with lower resistance, though.
The inductance doesn't play a role? With what resistance would you recomend to find a ferrite bead?
https://gr.mouser.com/c/passive-components/emi-filters-emi-suppression/ferrites/ferrite-beads/?package%20%2F%20case=0603%20%281608%20metric%29
-
It's not a switching converter, here the inductance simply acts as a filter for high frequency noise.
The higher the better, but keeping DC resistance low.
BLM18KG102SN1D (https://www.mouser.es/ProductDetail/Murata-Electronics/BLM18KG102SN1D?qs=0lQeLiL1qyar86AJwfwIYw%3D%3D) seems a good candidate, 1Amp rating, 0.1Ohms at DC but 1K impedance @100MHz.
Xc of 1.7uH at 100MHz is 1068 Ohms. So it seems to match.
Personally? I would leave as it is. It's a 500GB mechanical drive, probably has more than 10 years.
Between parts and shipping cost it'll be half of a new drive!
-
It's not a switching converter, here the inductance simply acts as a filter for high frequency noise.
The higher the better, but keeping DC resistance low.
BLM18KG102SN1D (https://www.mouser.es/ProductDetail/Murata-Electronics/BLM18KG102SN1D?qs=0lQeLiL1qyar86AJwfwIYw%3D%3D) seems a good candidate, 1Amp rating, 0.1Ohms at DC but 1K impedance @100MHz.
Xc of 1.7uH at 100MHz is 1068 Ohms. So it seems to match.
Personally? I would leave as it is. It's a 500GB mechanical drive, probably has more than 10 years.
Between parts and shipping cost it'll be half of a new drive!
New drive I can't find with same firmware so for that I repaired. Thanks you again for your precious help.
-
(https://i.imgur.com/pnNGwLi.png)
They are protection diodes. You can replace them with 0 ohm or jump with solder.
But before you do test for short circuit. If no short circuit proceed as above. If there is short circuit, find what is shorted. :popcorn:
You should give it a thorough inspection under microscope with multimeter in diode mode. Beware, if you screw up, you won't recover your data. So if it's important, send it reputable repair shop and pay money to get it fixed -- NorthridgeFix for example.
-
(https://i.imgur.com/pnNGwLi.png)
They are protection diodes. You can replace them with 0 ohm or jump with solder.
But before you do test for short circuit. If no short circuit proceed as above. If there is short circuit, find what is shorted. :popcorn:
You should give it a thorough inspection under microscope with multimeter in diode mode. Beware, if you screw up, you won't recover your data. So if it's important, send it reputable repair shop and pay money to get it fixed -- NorthridgeFix for example.
I repaired already and work fine.
-
New drive I can't find with same firmware so for that I repaired. Thanks you again for your precious help.
I mean a new drive - any new drive. Why it does to be the same drive and FW?
Is it for a special machine with license locked to the HDD? In that case, the money in repairing an old drive is totally justified.
-
New drive I can't find with same firmware so for that I repaired. Thanks you again for your precious help.
I mean a new drive - any new drive. Why it does to be the same drive and FW?
Is it for a special machine with license locked to the HDD? In that case, the money in repairing an old drive is totally justified.
Is a PC based machine and it for some reason it checks the disk firmware in order to live you to install the image.
-
Yeah, I suspected so when you kept saying it must had the same firmware.
I had similar repairs in the past, 2000s or even 90s machines which costed millions when new.
Thus they didn't care if a 128MB RAM stick was $100, they just wanted it back to service YESTERDAY :D.
-
Yeah, I suspected so when you kept saying it must had the same firmware.
I had similar repairs in the past, 2000s or even 90s machines which costed millions when new.
Thus they didn't care if a 128MB RAM stick was $100, they just wanted it back to service YESTERDAY :D.
You truly understand me.
-
I recommend finding a way around the software/firmware lock. Regardless if you manage to get the disk fixed.
-
I recommend finding a way around the software/firmware lock. Regardless if you manage to get the disk fixed.
Also,
I made a disk by replaced the SH6964B TI Motor Controller (tested and works fine).