Vintage harddisk sounds - My Seagate drives smaller than 1 GB 

Thanks for your answer and sharing your experience with the ST3243A. It's strange, that it lacks so much behind the other models of this drive family. The ST3390A is a completely different animal. I think this might have been one of the fastest drives back in the day together with the WDAC2340H.

Yes it has a very distinctive sound that is trademark for all of the Seagate drives with this design, more or less. Back then you could quite reliably tell the drive model that was installed in a PC by listening to the startup sound ^^

@@gennidee Absolutely. My ST3491A makes exactly the same noise. I always get goosebumps when hearing that distinct rattle. It was in my first ever PC i've owned (and still own), and the drives still works.

Vacuum cleaner

I ran a little file server myself just for my own purposes when I was in college. It was a beautiful HighScreen big tower full with Seagate and IBM 68pin SCSI drives. That thing was obnoxiously loud ^^

Each has its distinct sound (: very nostalgic.

Yes this is an odd drive. Very comparable in speed to its slimmer brother the ST3144A despite its size and it is a bit quieter if I remember correctly. Mine has one bad sector that seems to disappear once its heated up. Also mine randomly freezes the PC. Don't know what that's about. The Molex connector problem was not observed on mine. I think it was a cool drive back in the day.

Thanks for your response. I was hoping someone would comment on my question regarding the ST3243A (: Interesting that they made different versions of that although it might be just cosmetics or minor firmware updates because by the end of 1994 there were bigger drives around (400...800MB) and the ST3243A was entry level so I don't think they spent much resources on im proving it by then. By the way if you test this drive with SpeedSys it first starts with a "Linear verify" read test that is significantly higher on this (and other) drives then the actual "Linear read" speed test that comes next. Linear verify averages at 574kB/s with ZBR steps at about 40% and 80% (starts with ~700kB/s and ends with ~550kB/s). So these numbers are about the same as your drive. The linear read on the other hand starts at just about 200kB/s and is fairly constant from start to finish. I had several drives that showed vast differences between "linear verify" and linear read" speeds and the linear read is always consistent with the results of other benchmarks or a simple "large file copying onto empty drive" operation under DOS. Other drives show no differences between "linear read" and "... verify" though so I don't quite know what's going on. How is your ST3243A holding up with the "linear read"? About the ST1144A: I'm also very surprised how quiet this drive is (spindle motor and seek noise) compared to its low profile contemporaries. I know there is not much difference in the video but in person it is just a bit quieter than, say, the ST3144A. Pretty nice drive in my opinion. Mine has two dead sectors though which is unfortunate. Apart from that it's a solid performer for what it is. I am a big fan of the ST351A/X. Actually this model started my collection back in about 1995 or so when I was a kid and found a pretty beaten up 386SX big tower on the street that had a little fire on some of the wires. I took it home on my bicycle and took it apart and it had a ST351A/X in it that I still own to this day. Had a lot of drives since then that I don't own anymore but this one was always supposed to be a keeper. Much later, like 20 years later or so, I started searching eBay for ST351A/X drives and was pretty stunned on the prices. Sometimes I got lucky, though, and got one for cheap although cheap is a matter of perspective here. Three of the six I aquired from eBay over the last few years were in the 20...40€ price range which, compared to what I usually pay for old drives (generally 1-figure prices) is top dollar ^^ But well .. I like these as you could guess from the fact that I have two individual videos about this model and it is featured in four videos (: Keep looking for one, they are pretty nice :

@@gennidee The 640KB/s figure is the Read speed, Verify is indeed higher at 929KB/s. There is also a single spot where speed lowers and not two, so I guess despite having the same model number they have quite a significant difference. The graph looks like this i.imgur.com/gWFuNZS.jpg. I do have at least another drive having different read and verify speed, a Conner one which actually has a big difference between the two, so that's not specific to the ST3243A at least.

You're right, that is definitely a whole different drive. My suspicion is that this is a depopulated version of a bigger drive. Maybe they manufactured two identical drives and made a smaller one to fit a market niche or they used faulty platters by disabling parts of the surface. I think the fact that the access times are completely different might support this theory. I have two ST3243A and both are very much the same measuring 21ms average seek, 31ms maximum seek and 6.5ms track2track seek. Now check your results. Your drive is quite a bit faster, more of the likes of the ST3391A that you mentioned earlier. My ST3391A measures 15.6ms/22.1ms/4.6ms (avg/max/t2t) which looks a lot like it shares the same head stack like your ST3243A. The only thing that does not compare between your ST3243A and my ST3391A is the reading speed. My ST3391A measures equally for "linear verify" and "... read" and averages at 1.6MB/s. So who knows .... Here are the saved screenshots from SpeedSys: ST3391A imgur.com/a/oadNoHJ ST3243A imgur.com/a/PiileNT

​@@gennidee Well the numbers looks closer to the ST3391A with lower read speed (and obviously lower capacity) but I guess we will never know what the difference is for sure, I guess they could have improved the drive's firmware in a year too. It definitly doesn't have the same seektest as the ST3391A at least, and like yours it doesn't performs its full seektest unless plugged into an IDE controller. From the small seeking part that happens when there's just power, it sound almost identical to yours, mine just sounds a little bit lower pitched and it seems a little slower doing it but it probably doesn't mean much. I wonder if yours has the same full seektest when plugged into the IDE controller with the two "Clicks" at the end (As I didn't know how it would sound when I first powered it on, these two clicks were really worrying as usually a drive clicking isn't something good, but I did check it for bad sectors and there's none, so I assume these two clicks is just how it's seektest is supposed to be). Well, I guess you'll need to find someone with the same version as yours with the older label to see if they can replicate the slow read speed then.

By the way I went through my list and these drives in my collection do have way lower "linear read" speeds than "linear verify": Conner Peripherals CFS210A Conner Peripherals CP30254 Conner Peripherals CP30254H IBM DHAA-2405 Seagate ST3243A Western Digital WDC AC1170M It is worth mentioning that there are several drives where it is the other way around: "Linear verify" shows significantly lower values than "linear read". Dunno what the cause is for both variations.

Thank you for your comment. Both very common drives. I still see them occasionally on eBay auctions and they seem very reliable. None of my 3 ST3120A has any bad sectors, neither do the other Seagate drives with the same outer design. The only exception is the ST3243A where both drives I own are horrendously slow and one has several bad sectors. Very strange regarding the outstanding reliability of the other drives of this family.

I don't use those drives anymore since i now use CF cards for my retro PCs.

@@pentiummmx2294 Which is probably the better option since it's faster and easier to transfer data between PCs. Still I would miss the sound of those old drives which for me is part of the vintage experience. I make backups of my system drives with Norton Ghost though (: Also 95% of my drives are only collector's items. I don't really use them.

I remember using the st3144a on a micron 486 pc I had back in like 97.

On these older Seagate drives you can find the date stamped on the PCB. It has always the format YYWW and is written using white ink. So for instance mine said 9618, meaning 1996 and the 18th week. They eventually transitioned to printing the date code on the label. Don't know the size range or model when they changed it. There might be other numbers stamped on the PCB as well but usually it is pretty clear which number is the date code given the approximate time the drive must have been manufactured.

Don't you think 1996 is a little too far for 545.5MB? I have a Conner CFS1081A from 1995 and it is 1080MB.

@@cdos9186 I purchased a P150 with a 1.6gig Seagate/Conner drive in August 1996, so I know what you mean. But there was always a low end drive for cheap on the market. Compare for instance the WDAC2200 and the WDAC280 in my other video. The WDAC2200 is from December of 1992 while the WDAC280 is from March of 1993. Also the ST351A/X in this video is from 9236 (36th week 1992) and hence even younger than the ST3120A which dates 9213. So yeah I think it is completely possible for 545meg drives to have been produced in 1996 while 1.6gig drives were made in the same year. Probably to have a drive available in every price range: For retailers who sell cheap PCs and sellers of high end PCs. Both were able to find a drive in the Seagate portfolio. Just like today where new 500gig drives coexist with new 12TB drives.

The 12TB Seagate Exos has 8 platters! Also the "low-end" drives are not really low-end because 5 years ago those were in just about every PC you could buy at that time, they were the high-end drives. In 5 years the Seagate Exos 12TB is going to be a low-end drive. These drives are not really low-end and you getting a quality drive for cheaper because there is something better on the market. It works like how a high-end PC 15 years ago was $2200 and now they are considered low-end and you can get them for as little as $50 now because there is something better on the market. Take for example a new Dell will be considered low-end in 10 years and you can probably then get them for $110 instead of $1100. I keep all of my old computers and laptops and I don't basically throw away my money because 10 years ago I could have paid $850 for that computer and people basically just throw things away. Something low-end would be something that is made cheaply and a quality drive from 10 years ago isn't really low-end. Kinda like how even a 10 year old laptop would have been considered extremely high-end in 2000 because they didn't even have dual core processors in that time. If they did it would probably cost around $4000 for something like that if a 2009 laptop existed in 2000. So you are still getting a quality laptop if you buy something that was high-end for the time but it just isn't as powerful because of the technology used in that time.

Also you have some VERY rare drives in the video.