I bought 4 bad hard drives, can I make them work?!
HTML-код
- Опубликовано: 9 фев 2024
- I bought 4 hard drives listed as bad/non functional from 3 different sellers. Will any of them work? Can I fix them? Watch the video to find out!
The miniscribe 8425 was 20€ + 17€ shipping.
The Seagate ST-251 was 20€ + 12€ shipping.
The Seagate ST-157N & Quantum ProDrive ELS 85S lot ended at 8.27€ with 4.90€ shipping.
Yes, the first two weren’t actually that cheap for drives listed as bad.
I went for these as Miniscribe drives often have stuck steppers so I hoped it would be that, the ST-251 pictures seemed to show a pattern of the odd numbered heads being completely dead, and I thought it might be a head select signal not coming through, and the Quantum was likely to have the rubber issue and it along the ST-157N were cheap.
I have actually bought another lot of bad drives, a big one with a lot more drives, that ended up very cheap, 4€ + 22€ shipping for 13 drives, or basically 2€ per drive. They are newer drives than these though, they probably won’t be fixable but seeing the inside will be nice and who knows maybe some of them will do something. Expect a video on these when I get them!
The music at the start is an old song from NCS, Jim Yosef - Eclipse. • Jim Yosef - Eclipse | ... 
Наука
When I heard the ST157N spinning up, I almost fell off a desk chair
Yeah that sound was nasty, when I spun it up and heard that I wondered if one of the heads was ripped of, but clearly it wasn’t the case. Still that kind of heads rubbing on platters must not be very good for it.
Some answers: the seller listed that 251 as an "original" model, but it pretty clearly says MLC-2 in two different spots, meaning this is among the last made with the best performance from Seagate's low impedence stepper.
Also, that ST-157N not being parked initially is usually as a result of the drive being slammed around in transit.
I am also fairly sure that the Miniscribe 8425 only has 612 user accessible cylinders. You might want to try that.
A lot of drives are listed as bad simply because most people cannot or will not test them!
Yes it is an MLC-2. I have noticed the seektest is noticeably shorter than most of the ST-251 videos on RUclips, and when I ran spinrite I was surprised by how fast the seek times were. Stason/TH99 doesn’t have details for the MLC-2 but the seek times appeared to match those that they state for the ST-251-1.
I wonder if there’s a place to find the differences between all those variants as apparently there’s at least ST-251 0, 1, MLC-0, MLC-1, MLC-2 and possibly more.
The side of the drive has a stamp that says x8751, so made very late 1987. I have no idea when the ST-251 started and stopped being made, I have heard of people having some made up to the early 90s though so it must not be that late?
But it’s definitely not the original slower one, I don’t know why the seller thought that was the case. The ST-251 label itself is smaller and doesn’t have the font than the early ones have.
I’m not sure what’s the deal with the ST-157N, I didn’t even think shipping could really make a stepper motor move, and if the shipping was so bad, it probably would’ve damaged the surfaces wouldn’t it? I wont question it too much, it just was strange.
The ST-251-0 is the slowest model of the series while the ST-251-1 was the faster and more reliable version. MLC is Manufacturing Line Code. So even if it had MLC-0, MLC-1, or MLC-2, the performance is the only thing separating them. The ST-251 was introduced in 1990 and they're known for their hollow knocking noise on some of them due a misaligned track 0 sensor. Also, ouch to that squeal on the ST-157N.
As for the Miniscribe 8425, were you able to fix the PCB on it?
@@Kali_Krause I have seen the slower ST-251 on a single video and the top just said ST-251 with a label that looked the same as what is on the ST-225. The seektest of it sounded significantly different compared to the others ST-251 I’ve heard including mine. What’s sure is my MLC-2 here has a seektest that’s shorter than the more common -1. The date code on it indicates it is from very late 1987, but I have seen pictures of ST-251 with a date code from 1986. As far as I understand the track 0 detection is from some kind of positioning data on the platters, if it swipes the entire surface on the slow seek part at the start of the seek test and knocks a few times, it means the drive could not find that positioning information. I’ve seen at least one video where the drive did exactly that and worked anyways, but yeah it seems like getting the full swipe with knock at the end means something is going wrong.
Yeah the ST-157N sound was really nasty but it seems to be working great surprisingly!
No the miniscribe still doesn’t work, and I’m not even sure if it’s the PCB as I have been told a bad head could cause that as well. I have opened it and didn’t see a obvious issue in there but yeah. Maybe one day I’ll eventually get another one and will be able to know for sure which part is the one that’s bad
I hope you get that miniscibe to work, I have a similar 8225 and it sounds great
I do not think that will happen, but maybe one day I’ll get another one and be able to check if it really is the PCB that is the issue (cause apparently it could also be a bad head that causes this)
Very cool!
Yes indeed! Ending up with 3 drives that work out of 4 is pretty good. Two of them with no bad sectors is even better! I wish I could’ve fixed the Miniscribe too but I really can’t complain considering the circumstances.
of course you made them work. every drive that you buy comes back to life
I wish that was true lol. The very big pile of platters and magnets I pulled out of dead drives says otherwise
Would you recommend use a low level format for a old HDD?
I have a 4.2gb Maxtor 4320 with 26 reallocated sectors, the rest of the drive is apparently in very good condition.
Only MFM drives, and (maybe) very early IDE drives can take a low level format. (SCSI drives can be formatted from the controller but it isn’t possible to really know if they are doing a true LLF or not)
Modern "Low level format" tools meant for more recent drives will just write the entire surface with 0s.
Doing a full surface write can be a good idea as it might help discover more bad sectors if they exist. Once the sectors are reallocated they will stay that way.
I, for one, got a Bigfoot 2160AT with quite sad results - it will complete seek test, but that's it. Trying to access it results in sounds similar as if the head is stuck, though a quick look shows the head moving, trying to read the sector over and over again.
I'll have another lot coming through, comprised of:
- Quantum Bigfoot CY 6,4GB (hope this one works!)
- Fujitsu MPB3021AT-E 2,16GB
- Seagate Medalist 4321 4,6GB
- Fujitsu MPF3102AH-ST 10.2GB
- Quantum Fireball Plus AS 20.5GB
- Quantum Fireball 635AT 630MB (also hoping this one works as I have a 486 that needs a small sized HDD)
I'll try to get around setting up a PC to record all the drives I have (aside from my dual P3 Coppermine machine which is probably the first to be recorded - WD740 Raptor over SATA2IDE adapter + ST336607LW SCSI Seagate Cheetah) and hopefully make a compilation of them all.
Drives like that Bigfoot are the worst, acting like they might work but yet they don’t… I have had some that did similar things to that.
That seems like a nice lot of drives, hopefully they will work!
@@arnlol They arrived today.
So far, testing ended with the MPF3102 being dead (fails seektest), and the 635AT seems to stop reading after 310MB or so - I suspect some developed bad sectors on the 635AT which might be recoverable.
The Bigfoot turned out to be another 2160. All drives except the 635AT (which will hopefully survive Spinrite or DiskGenius) and the MPF3102 have passed with flying colors, and as a bonus, the Fireball Plus AS came with a 98SE install (that I unfortunately could not boot on the dual CPU machine due to RAM being too much lol).
The Medalist is surpisingly silent (beats my other 4.3GB drive, an IBM from ~1997 or so), and nearly the same can be said about the Bigfoot and Fireballs, as well as the working MPB3021.
As a last resort I might clean up the contacts on the MPF3102 and give it another go - it did look quite dirty compared to all the other drives, and from my previous experiences (usually with newer WDs but not limited to them), cleaning contacts always helped.
I might explore the 20GB Fireball soon to see what's in there (though it's probably just a bare 98SE install, given the rest of the working drives were wiped clean.) soon - just have to plug up my 440LX board (QDI Legend V/P6I440LX) and it should boot without issues.
@@NSHG Sounds like it went decently well.
Do you mean the spindles are silent? It's always good to get drives with good bearings.
I'm not sure how the original Fireballs react to bad sectors, I have two but both are working properly. Maybe there's a small bad area that you can format around, or worst case make a 300MB partition hopeing it doesn't get worse over time.
Can the kind of connector like the one on the Fujitsu really be affected by dirty contacts? I guess it doesn't hurt to try anyways.
@@arnlol Yeah, most of their bearings sound quite well for their age - especially the 635AT I specifically bought the lot for, given it'll serve as a main drive for a 486 DX2.
I'll check if the 635AT might have recoverable space from reallocation - there's no SMART on it! 😂
As for the Fujitsu, I'm suspecting some of the contacts on the ribbons being dirty. Not having high hopes for it but worth a shot, as I previously had some 80GB WDs do funny stuff due to oxidation.
nice
Wow was not expecting most of those to be working being listed as bad! Also for the 8425, are you able to use HDAT to seek the drive around? Perhaps there is indeed a issue with the PCB or heads but maybe you just have the jumpers set wrong? That is just odd to have it not even try to format and get stuck like that, I've had that happen when the drive jumper was set wrong myself but not sure about that one. HDAT will definitely at least make the drive try and seek around so would be curious to hear if that did anything. Also how in the world did you install Windows 95 on that ST-251??? Don't you have to disable onboard IDE which means you cannot boot from the installer CD to install Windows 95? I am baffled by how you were able to do that!
The jumpers are correct on the 8425, it just reports a write fault (which it wouldn’t if the jumpers were wrong). Yes it does seek (doesn’t need to use HDAT either, like speedsys seek part works on it), it just won’t write anything at all always giving a write fault.
As for 95 on the ST-251 used the floppy version of 95 RTM, and ran setup /id to bypass the disk space check. For some reason setup says it will take 43MB and there’s like 41MB free, but it totally works and ends up with almost 10MB free at the end…
@@arnlol Sounds like the 8425 probably has damaged heads? It seems like a common fault on those for some reason, I have seen multiple that act that same way and refuse to take a format. Although you could be right about the PCB having an issue, the only way to really know is to have a known working one and swap over the PCBs and if it still doesn't work with a known good PCB it definitely has a few or multiple dead head signals.
Interesting way of installing Windows 95 on that ST-251! Do you have a set of the factory floppy disks or did you just image them from online? I think that is something like 17 or 19 floppies to install that and I assume that takes a lot of time doesn't it?
@@cdos9186 VCS told me on discord it might be from a bad head, I opened the drive and didn't noticed ripped off or visibly damaged heads but who knows maybe it's a bad head... How am I supposed to know... Maybe one day I'll find another one and will be able to test. (I assume I could swap PCBs with the SCSI varients?)
No I don't have a set of factory floppies, I made those months ago (years at this point?) from images, there's 13 floppies. It took about 30 min to install.
@@arnlol is it possible to probe the signals coming from the heads or even better...
since the MFM boards on the hard drives are very "dumb" isn't there a software to do hardware level analysis? like to see if the drive is ready, if the track zero was found, to raw read what is passing under the heads etc?
je ne savais pas que le plus grand collectionneur de HDD étais en France !
Oui je suis bien Français mais par coutre le plus grand collectionneur, oulaaa non.
Déjà rien que la personne qui gère la chaine DiskMaster il a plus de disques et en plus ils sont plus anciens, des années 80s donc plus chers/bien plus durs a obtenir. Il y a aussi un site old-harddrives.net, ou la personne as apparemment le Guinness world record de la collection avec le plus grand nombre de disques durs différents. Moi j'en est seulement aux alentours des 250 et il y a quelques doublons, en plus la majorité sont des disques des années 90 et début 2000, mais bon je fait ce que je peux avec le peu d'argent que j'ai, en vrai c'est déjà pas si mal :)
@@arnlol c une hyperbole mais c un bonne petite collection que vous avez ! Mon plus vieux disque c un wd800jd-23jnc0
@@le_mody335 Mon oncle a exactement le même dynamisme !
2:23 - 2:25 was horrible! not a sound you want to hear Turing a computer off 😂
I think the lubricant that’s meant to be on the platters is probably worn off hence the noise, but that drive didn’t recover anyways. Yeah you probably don’t want to hear that
Bad drives aren't bad if you don't try to repair them first!
Most people will not try that though. Buying bad drives isn’t a good idea but yes sometimes they can be fixed, and made "good" again.
Oh hi
Can't believe that I'm first
Well that’s not too hard when the videos only get a few hundred views to be fair x)