[SOLVED] How to repair old Samsung SP1604N HDD? It does not spin.

[tmoi]

Hi there guys, i'm new here.

So I found an old 3.5 HDD of mine and i know it has a lot of childhood memories on it.
For reference, it looks like this:

So the thing is, when i plug it into a power supply it does not start spinning. It does not even make a sound, so i guess its not stuck or anything, because then i would expect it to make stuttering sounds.

I know there can be lots of stuff causing it not to move, but i would think it should be one of two causes here:

1. The motor died.
2. The PCB died.

So I guess if its option 1, then i would have to do a platter swap to an exact same HDD, which i would fail due to dust and platter misalignment and all.
Thats why im focussing on option 2.

I read that for newer HDDs you can not just get an exact same PCB and mount it to the existing HDD.
But this HDD is from around 2003.

So the question is, can I just buy the exact same HDD again and mount the new PCB to my old HDD?

 

[tmoi]

Hello @Themx86 and thank you very much for your reply.

I can confirm that both my HDDs, the new one and the old one are the same model SP1604N and both have written PALO on them.

To sum this up:

1. The patient drive did not make a sound at all in the beginning
2. I put the donor PCB (different firmware) on the patient, since then the patient tries to spin up but aborts with one or two clicking sounds.
3. I put the donor heads onto the patient as well. The patient now has the new PCB and the new heads, but behaves the same.

Two things you mention are pretty interesting:

The firmware of the board should match the firmware of the heads.

I do have a match between these two at the moment.

I already tried swapping them three times without a tool, but using a custom made spacer instead.

I also built a custom spacer for the head swap. I can confirm that I saw none of the four heads touch each other during the swap. Who knows though, I might have missed a slight touch after all.

PS did you read your terminal output yet?

Sadly, no. Even with the instructions linked above this is way beyond my skill level. I know nothing about electrical circuits and all.

So after all, if the new heads are failing now and it sounds the same as with the old heads, then the original HDD had two problems it seems: A broken PCB and failing heads as well.

Oh and I will be waiting for your reply here. I hope you will succeed in doing a head swap with your tool then Let me know.

 

[tmoi]

@Themx86 I got some serious news.

I ordered another donor HDD two weeks ago and it just arrived.

So out of curiosity i took the new donor PCB, mounted it to the patient.

And the drive spins up and keeps spinning!!

Thats great, step one done.

Now there are two symptoms: Once every second the drive clicks when connected to the IDE cable. When removing the IDE cable the drive runs silent. So most likely it clicks when trying to read information, which would be failing heads, right?

Also the drive is not recognizes by windows. Sometimes i see it appear and getting a drive letter, but even then it says i can not access this drive.

Any inputs on how to proceed from here? Thank you very much.

 

[Themx86]

@Themx86 I got some serious news.

I ordered another donor HDD two weeks ago and it just arrived.

So out of curiosity i took the new donor PCB, mounted it to the patient.

And the drive spins up and keeps spinning!!

Thats great, step one done.

Now there are two symptoms: Once every second the drive clicks when connected to the IDE cable. When removing the IDE cable the drive runs silent. So most likely it clicks when trying to read information, which would be failing heads, right?

Also the drive is not recognizes by windows. Sometimes i see it appear and getting a drive letter, but even then it says i can not access this drive.

Any inputs on how to proceed from here? Thank you very much.

The following could also be possible unfortunately,

Your PCB

Hello @Themx86 and thank you very much for your reply.

I can confirm that both my HDDs, the new one and the old one are the same model SP1604N and both have written PALO on them.

To sum this up:

1. The patient drive did not make a sound at all in the beginning
2. I put the donor PCB (different firmware) on the patient, since then the patient tries to spin up but aborts with one or two clicking sounds.
3. I put the donor heads onto the patient as well. The patient now has the new PCB and the new heads, but behaves the same.

Two things you mention are pretty interesting:


I do have a match between these two at the moment.



I also built a custom spacer for the head swap. I can confirm that I saw none of the four heads touch each other during the swap. Who knows though, I might have missed a slight touch after all.



Sadly, no. Even with the instructions linked above this is way beyond my skill level. I know nothing about electrical circuits and all.

So after all, if the new heads are failing now and it sounds the same as with the old heads, then the original HDD had two problems it seems: A broken PCB and failing heads as well.

Oh and I will be waiting for your reply here. I hope you will succeed in doing a head swap with your tool then Let me know.

Well there are two scenoarios also possible

A - Your PCB was broken, and due to the wrong firmware on that moment, the harddrive starts clicking. You replaced the heads with the same firmware, but the heads where broken in the mainwhile.

B - Your PCB was broken, and malfunctioned due to firmware issues. You actually succeed in replacing the heads, but it cannot find the first sector, because it has a different location due to firmware incompatibilities.

The clicking is basically it cannot find the first sectors to read from. In these sectors the calibration of the hard-disk is stored (which normally is in the special ROM chip on the PCB).

In both scenarios, you should replace the PCB and the HEAD with the original Firmware version. If i read correctly 100-24, otherwise it will never find the first sector, with or a broken head or firmware that does not match the first sector.

Its a ** drive to repair. May I ask in which country your located?

Greets

 

[Themx86]

@Themx86 I got some serious news.

I ordered another donor HDD two weeks ago and it just arrived.

So out of curiosity i took the new donor PCB, mounted it to the patient.

And the drive spins up and keeps spinning!!

Thats great, step one done.

Now there are two symptoms: Once every second the drive clicks when connected to the IDE cable. When removing the IDE cable the drive runs silent. So most likely it clicks when trying to read information, which would be failing heads, right?

Also the drive is not recognizes by windows. Sometimes i see it appear and getting a drive letter, but even then it says i can not access this drive.

Any inputs on how to proceed from here? Thank you very much.

Wow thats a good step. Can you confirm that the PCB firmware is the same as your old original board? Thats great since the firmware would only be stored on the PCB and the First sectors!

For your question could be failing heads, properly bad sectors.

there are some techniques to read around the bad sectors.

Do you use linux, mac or windows?

 

[tmoi]

Wow thats a good step. Can you confirm that the PCB firmware is the same as your old original board? Thats great since the firmware would only be stored on the PCB and the First sectors!

For your question could be failing heads, properly bad sectors.

there are some techniques to read around the bad sectors.

Do you use linux, mac or windows?

Thank you so much for taking the time to reply.

I forgot to mention before that the second donor drive comes with the same firmware as my original patient drive.

So the setup that spins and keeps running is:

• platters and motor from my original firmware 100-24 drive
• heads from donor 1 drive which was firmware 100-30
• PCB from donor 2 drive which again was firmware 100-24

So yes, the original PCB and the current PCB both have the same firmware.

It seems like the heads do not really know or care about the firmware then, right?

The clicking which occurs now ever second sounds very different from the clicks that happened with the donor 1 PCB, aka wrong firmware PCB.

I'm not quite sure where to go from here. Should I do another head swap to get the heads from donor 2 drive into my patient, so that the heads also come from a 100-24 drive? I'm really not sure if the heads care about the firmware.

For your questions: I'm running windows and I am located in Germany.

Again, thank you so much for your replies.

 

[Themx86]

Thank you so much for taking the time to reply.

I forgot to mention before that the second donor drive comes with the same firmware as my original patient drive.

So the setup that spins and keeps running is:

• platters and motor from my original firmware 100-24 drive
• heads from donor 1 drive which was firmware 100-30
• PCB from donor 2 drive which again was firmware 100-24

So yes, the original PCB and the current PCB both have the same firmware.

It seems like the heads do not really know or care about the firmware then, right?

The clicking which occurs now ever second sounds very different from the clicks that happened with the donor 1 PCB, aka wrong firmware PCB.

I'm not quite sure where to go from here. Should I do another head swap to get the heads from donor 2 drive into my patient, so that the heads also come from a 100-24 drive? I'm really not sure if the heads care about the firmware.

For your questions: I'm running windows and I am located in Germany.

Again, thank you so much for your replies.

Ok good to know,

Yes so firmware is only stored in the PCB and first disk sectors.

The ticking you hear is properly a bad sector it wants to read (due to dust particles while opening the drive)
Head swap is your last resort now.

If your BIOS/Computer recognizes the drive and considering the drive is a bit damaged, you should turn your drive off, preventing it for causing more damage.

Do you have the knowledge or computer available with UNIX and a hard drive with the same or bigger capacity?


Well i'm located in the Netherlands

 

[tmoi]

Ok good to know,

Yes so firmware is only stored in the PCB and first disk sectors.

The ticking you hear is properly a bad sector it wants to read (due to dust particles while opening the drive)
Head swap is your last resort now.

If your BIOS/Computer recognizes the drive and considering the drive is a bit damaged, you should turn your drive off, preventing it for causing more damage.

Do you have the knowledge or computer available with UNIX and a hard drive with the same or bigger capacity?


Well i'm located in the Netherlands

Hi neighbor I guess

Sadly I do not have access to any unix system or any bigger drive like this one.

So if the drive suffers from physical damage, like the dust you mentioned or even scratches that any failing heads might have caused, then I guess there is not much I can do right?

You said a head swap is my last resort. It still wouldn't help if the drive has physical damage though right?

The computer recognizes the drive and assigns a drive letter. But it won't let me access it.

You said in that case I should turn it off to prevent further damage. So there are no options for me to try to access the data without a unix system and a bigger HDD? Man I'm screwed. I get the feeling that I should have waited with the execution of the first head swap until the donor 2 with matching firmware arrived. But at that time I thought I would never find one matching as these drives are not that common anymore.

So I guess I will attempt another head swap to get the matching heads into the patient?

Thank you.

 

[Themx86]

Hi neighbor I guess

Sadly I do not have access to any unix system or any bigger drive like this one.

So if the drive suffers from physical damage, like the dust you mentioned or even scratches that any failing heads might have caused, then I guess there is not much I can do right?

You said a head swap is my last resort. It still wouldn't help if the drive has physical damage though right?

The computer recognizes the drive and assigns a drive letter. But it won't let me access it.

You said in that case I should turn it off to prevent further damage. So there are no options for me to try to access the data without a unix system and a bigger HDD? Man I'm screwed. I get the feeling that I should have waited with the execution of the first head swap until the donor 2 with matching firmware arrived. But at that time I thought I would never find one matching as these drives are not that common anymore.

So I guess I will attempt another head swap to get the matching heads into the patient?

Thank you.

I would recommend you to not do any head swap before your tried to backup the drive first. It also really depends on what your looking for. Properly you have a few damaged sectors, and if for instance you are looking for photos, you can recover 99% of your photos. If your drive has three movies of 50 GB each, you will properly loose a part of every movie

I highly doubt a head swap will improve things for you, only if one of the four heads are not being used to read the first part of the drive, but you can only be sure when trying software, on Linux, while backing the drive up to a bigger one. Every time you open up the drive, more sectors become damaged, it even happens in clean rooms.

All depends on the data. I get for instance a lot of clients with old bitcoin wallets. a wallet is around 128KB - 3 MB. In this particular case the change of recovering is almost 99%, since it a very small change its exactly on the damaged sector.

 

[tmoi]

I would recommend you to not do any head swap before your tried to backup the drive first. It also really depends on what your looking for. Properly you have a few damaged sectors, and if for instance you are looking for photos, you can recover 99% of your photos. If your drive has three movies of 50 GB each, you will properly loose a part of every movie

I highly doubt a head swap will improve things for you, only if one of the four heads are not being used to read the first part of the drive, but you can only be sure when trying software, on Linux, while backing the drive up to a bigger one. Every time you open up the drive, more sectors become damaged, it even happens in clean rooms.

All depends on the data. I get for instance a lot of clients with old bitcoin wallets. a wallet is around 128KB - 3 MB. In this particular case the change of recovering is almost 99%, since it a very small change its exactly on the damaged sector.

I only want to get small data like photos back If I can. That's good.

But how would I be able to do a backup of the drive to a bigger one if the drive can't be read? Assuming I would get access to linux and a bigger drive anyway.

Also I do not know if the drive was formatted back then. If I create a backup, would I still be able to do data recovery of possibly formatted files as I would on the original drive?

We do not know where that first parts of the drive are stored right? There are two platters in the drive, is the first part on the top or bottom platter?

I'm glad you give me your input, this really helps.

 

[Themx86]

I only want to get small data like photos back If I can. That's good.

But how would I be able to do a backup of the drive to a bigger one if the drive can't be read? Assuming I would get access to linux and a bigger drive anyway.

We do not know where that first parts of the drive are stored right? There are two platters in the drive, is the first part on the top or bottom platter?

I'm glad you give me your input, this really helps.

The reason you can't read it with windows is because its trying to open its file system. In linux you can read it raw, clone it without the damaged sectors and copy it. you can use ddrescue for this.

attach the two drives, and enter:

ddrescue /dev/sda /dev/sdb

where sda is your broken drive
sdb is your damaged drive

 

[tmoi]

@Themx86 Sorry i missed your Post yesterday.

I will try to get access to a linux boot CD and boot my pc from it. Then I can try your suggested steps with ddrescue I guess? I have an external HDD which is big enough to host a raw copy of the patient drive.

As I missed your post, meanwhile I did some more research and testing, here are my findings:

The patient platters with donor 1 heads and donor 2 PCB spins up an keeps running as I mentioned, but it keeps clicking every second. Sometimes windows detects the drive but won't let me access it. Now I fired up HDD Raw Copy Tool from HDDGuru to attempt a raw copy of the patient drive. The tool detects the drive with its correct size and even starts the copy process. At some sectors it tells me there are irreparable read errors, but it just skips them and moves on. At 2% of the copy process it aborts though and tells me the drive was disconnected.

Still, I now have a 2% backup on my external drive, so I thought why not try data recovery software on this tiny bit of data. And indeed, the recovery software finds some data which was on the original drive and it also shows me some of the previous windows xp file structure, like my former user folder, desktop folder and documents and settings folder. Most of them are empty though, as the backup is only 2% of the original drive.

My questions about this: if the heads in this drive were broken / bad because I had damaged them during the head swap, would this partial recovery even have been possible?

Then I did another test. I assembled donor 1 again, which has the heads from the patient drive inside them. I know for a fact that these heads are bad by now because for learning purposes I let them touch each other before mounting them into the drive. Now the interesting thing is that this donor 1 drive spins up and keeps spinning, and it sounds exactly like the patient drive, clicking every second and not being accessible. So this makes me think the heads that I put into the patient drive got damaged during the head swap as well because the clicking is the same?

You say my chances of generating a full raw backup are much higher when using linux and ddrescue right? I will try this tonight.

As I'm not good with linux knowledge, can you link me to a linux boot CD which has the necessary ddrescue tool on it maybe? Thanks my friend.

 

[fzabkar]

HDDSuperClone is another Linux tool that understands how to clone a drive with bad sectors. The author has taken special effort to fine tune it for HDD applications. He tries to detect a bad head and skip over it.

Just FYI, the demo version of SHT/SHTR can dump the firmware modules from your drive:

http://www.hddoracle.com/viewforum.php?f=149

 

[tmoi]

@fzabkar @Themx86

Thank you for your inputs.

I am very cautious in what i try on this patient drive, as you guys told me i should really not power it on too often. I always keep that in mind now.

Still in order to create a copy of the drive, i got a bootable linux today and tried ddrescue on it.

The problem here is, that the drive hardly gets listed in the drive list ( I was using fdisk -l command in the linux terminal).

Most times the drive does not appear. Sometimes it does appear but immediately dissapears. One time it appeared in the list and i ran the ddrescure command to start the copy process, but then the drive disappeared again.

I guess also HDDSuperClone would not be able to find the drive if linux itself cant find it.

I feel like this drive is dying more and more everytime i have to power it on. The clicking sounds become a little more irregular and change in tone, volume and pattern every now and then now.

Any more thoughts? Thanks guys.

 

[fzabkar]

I'd be worried that a head may be scraping the platter. You can't get data from dust. :-(

It's too bad that you weren't able to retrieve a terminal report before you starting your invasive troubleshooting. :-(

 

[tmoi]

I'd be worried that a head may be scraping the platter. You can't get data from dust. :-(

It's too bad that you weren't able to retrieve a terminal report before you starting your invasive troubleshooting. :-(

I am pretty sure the heads do not touch the platters. I heard that on a different drive before and on my opinion the clicks from my current drive are way less 'destroying'. But I can't know for sure.

Regarding the terminal. I agree. The problem is I do not understand how to do this. I have zero knowledge about electrical components and could not understand the instructions you linked me to. If you manage to teach me, I would be super happy to get this terminal report.

 

[USAFRet]

I am pretty sure the heads do not touch the platters. I heard that on a different drive before and on my opinion the clicks from my current drive are way less 'destroying'. But I can't know for sure.

"I succeeded to do a head swap and I swear I did not damage the new heads at all. I was super careful and prevented the heads from touching each other or any platter during the whole process. "

The clearance between head and platter is on the order of 3 nanometers.
Much smaller than a spec of dust.

 

[tmoi]

"I succeeded to do a head swap and I swear I did not damage the new heads at all. I was super careful and prevented the heads from touching each other or any platter during the whole process. "

The clearance between head and platter is on the order of 3 nanometers.
Much smaller than a spec of dust.

Ok the way you phrased this kind of makes me realize I might have failed in doing a head swap. Damn.

 

[fzabkar]

You need to purchase a USB-serial adapter (approx. US$5). Then connect the adapter's Tx/Rx pins to the drive's Rx/Tx pins.

In order to see the terminal output, you would need a program such as HyperTerminal or PuTTY. A bad head would normally generate "gray errors".

 

[tmoi]

You need to purchase a USB-serial adapter (approx. US$5). Then connect the adapter's Tx/Rx pins to the drive's Rx/Tx pins.

In order to see the terminal output, you would need a program such as HyperTerminal or PuTTY. A bad head would normally generate "gray errors".

Ok. But how do I connect the adapter to the TX RX on the drive? Soldering?

Could you link me to an adapter? I think the ones I find are wrong.

 

[fzabkar]

Your drive should have a jumper block. The Tx/Rx pins are defined in the HDD Oracle tutorial.

This additional tutorial should help you:
https://www.instructables.com/id/Usb-to-SerialTTL-adapter/

 

[tmoi]

@fzabkar

Um so i have read both tutorials again, and here is what i understood:

The TX and RX pins are just two defined pins in my jumper block. I connect an adapter to them using cables.
Here is an adapter i found that also has the cables:

https://www.amazon.de/USB-TTL-Konverter-Modul-mit-eingebautem-CP2102/dp/B00AFRXKFU

It says the TX and RX are 5V. Is that fine or do i need 3.3V?

Is this adapter ok?

I do not have any possibility to do soldering at the moment, i hope i can continue without that?

If i connect the HDD to my adapter, how do i proceed to read the terminal report?

Thank you very much.

 

[fzabkar]

That adapter should be OK. The Tx/Rx signal levels would be 3.3V, not 5V.

If your drive has a jumper block, then no soldering is necessary. The tutorial explains what to do.