Yeah, I'm not able to find a download for the latest firmware. This lists them, though I've no idea which is supposed to be the latest, and there's no link to download any of them.

https://smarthdd.com/database/SAMSUNG-SP0842N/

The device manager just lists the controller/hard drive as a generic controller device.

DustyShinigami wrote on 2025-09-09, 21:02:

douglar wrote on 2025-09-09, 20:33:

Seems to me like the storage device is saying is that it's fine, but the controller is telling the operating system that there a […]
Show full quote

DustyShinigami wrote on 2025-09-09, 19:44:

These are the SMART details from MHDD:

The attachment IMG_4549.JPG is no longer available

Seems to me like the storage device is saying is that it's fine, but the controller is telling the operating system that there are problems. The root cause is likely one of these:

1. failing IDE controller in the system
2. cables not working at the speed negotiate between the controller and the storage device
3. an incompatibility between the controller device driver & the storage device firmware
4. edge case system memory that works fine with the CPU but fails intermittently on DMA requests

Hmm. Curious. I hope it's not the first possibility. Could using an 80 cable with a male to female adapter be causing an issue with the speed? I can't seem to find any 80 cables that are exactly 40 pins, only 39, so I'm not able to connect it completely from drive to bus.

Not sure if I already have the latest driver installed for the controller. 1.25 was what I installed. I'll have to see if I can find the latest firmware for the HDD. I take it nothing can be done if it's failing intermittently?

That missing pin is pin 20 and on the vast majority of hard drives and controllers is not used for anything other than being a missing pin used as a key to make sure you have the cable oriented correctly. Some rare controllers will use pin 20 to pass +5v power to cf devices , but that is pretty rare.

Could you post a picture of what you are doing? I’m unclear about why you need a gender changer.

douglar wrote on 2025-09-09, 22:44:

DustyShinigami wrote on 2025-09-09, 21:02:

douglar wrote on 2025-09-09, 20:33:

Seems to me like the storage device is saying is that it's fine, but the controller is telling the operating system that there a […]
Show full quote

Seems to me like the storage device is saying is that it's fine, but the controller is telling the operating system that there are problems. The root cause is likely one of these:

1. failing IDE controller in the system
2. cables not working at the speed negotiate between the controller and the storage device
3. an incompatibility between the controller device driver & the storage device firmware
4. edge case system memory that works fine with the CPU but fails intermittently on DMA requests

Hmm. Curious. I hope it's not the first possibility. Could using an 80 cable with a male to female adapter be causing an issue with the speed? I can't seem to find any 80 cables that are exactly 40 pins, only 39, so I'm not able to connect it completely from drive to bus.

Not sure if I already have the latest driver installed for the controller. 1.25 was what I installed. I'll have to see if I can find the latest firmware for the HDD. I take it nothing can be done if it's failing intermittently?

That missing pin is pin 20 and on the vast majority of hard drives and controllers is not used for anything other than being a missing pin used as a key to make sure you have the cable oriented correctly. Some rare controllers will use pin 20 to pass +5v power to cf devices , but that is pretty rare.

Could you post a picture of what you are doing? I’m unclear about why you need a gender changer.

'A gender changer'... 🤣

Hope these are clear, but this is essentially my 80 yellow IDE connector going from the top HDD (E) into IDE 1 and shows the adapter.

DustyShinigami wrote on 2025-09-09, 19:44:

These are the SMART details from MHDD:

The attachment IMG_4549.JPG is no longer available

That looks fine. It is not a 100% guarantee it is, but very likely your issues are caused by something else. Curiously enough there are no interface errors here either, usually a flaky cable or controller would cause some.

Probably the most sensible way to troubleshoot at this point is testing the HDD in another computer.

Also - run memtest, just to be sure. Memory errors can often break things in unexpected and spectacular ways.

Archer57 wrote on 2025-09-10, 03:27:

That looks fine. It is not a 100% guarantee it is, but very likely your issues are caused by something else. Curiously enough th […]
Show full quote

DustyShinigami wrote on 2025-09-09, 19:44:

These are the SMART details from MHDD:

The attachment IMG_4549.JPG is no longer available

That looks fine. It is not a 100% guarantee it is, but very likely your issues are caused by something else. Curiously enough there are no interface errors here either, usually a flaky cable or controller would cause some.

Probably the most sensible way to troubleshoot at this point is testing the HDD in another computer.

Also - run memtest, just to be sure. Memory errors can often break things in unexpected and spectacular ways.

Okay. Well, I won’t be able to do anything with it now for a couple of weeks as I’m going away. Once I’m back I’ll give memtest a run and report back. Last I used the machine, the Scan Disk error disappeared. Not sure what I did differently. So far it hasn’t been an issue. Though watch it kick off again when I come back to it. 😅

A question I do have: I take it IDE 1 and 2 only support UDMA 2 and no higher?

It’s a shame I can’t make use of UDMA 4 and above like the other HDDs. 😕 Although the drive is only being used to store CD images. Though I take it the reading of the images would benefit from the increased speed?

Thinking about it, if I get one of those Ultra66 IDE PCI cards, that would probably solve the speed issue, if the drive is funny with the IDE 3 and 4 buses.

It is called ATA33 in the manual, which should be up to 33mb/s and more than enough for a system and use case like this.

If you want to you can get newer cards too, all the way to ATA133 or even SATA. I have w98 on sata ssd connected to such controller with 1xIDE and 1xSATA port. Newer controllers have less issues with newer HDDs too...

But IMO, as i said, it is fine as is.

Archer57 wrote on 2025-09-10, 09:28:

It is called ATA33 in the manual, which should be up to 33mb/s and more than enough for a system and use case like this.

If you want to you can get newer cards too, all the way to ATA133 or even SATA. I have w98 on sata ssd connected to such controller with 1xIDE and 1xSATA port. Newer controllers have less issues with newer HDDs too...

But IMO, as i said, it is fine as is.

I was thinking of getting one of those IDE to SATA connectors. If I get an ATA133 card, I just want to make sure it has a couple of those LED connectors on it so I can hook my other two HDDs to it. 🙂

DustyShinigami wrote on 2025-09-09, 23:25:

Hope these are clear, but this is essentially my 80 yellow IDE connector going from the top HDD (E) into IDE 1 and shows the adapter.

Thanks for the pictures. 4 things are going against you there--

1. Mixing the 40 pin cable and the 80 pin cable is not a good idea. The 80 pin cable will tell your device that faster than 33Mhz is OK, but the 40 pin portion of the cable is unlikely to be able to handle that.
2. The max length of IDE cables is supposed to be <= 18" (45cm). If you go over 18", the chance of signal problems increases.
3. When you use an extension cable or an adapter in the IDE cabling, it creates additional signal reflections, and the chance of signal problems increases.
4. When IDE cables are running along side power cables , it can create additional signal challenges, and the chance of signal problems increases if you already have other challenges.

So the moral of this story:

1. If you use a single 80 conductor cable that is <= 18" and don't run it along side power cables, it will probably work fine.
2. With your current situation, maybe it works if you can force PIO 3, UDMA 0 or MWDMA 1. I'm not surprised that UDMA2 or faster has issues.

douglar wrote on 2025-09-10, 20:16:

Thanks for the pictures. 4 things are going against you there-- […]
Show full quote

DustyShinigami wrote on 2025-09-09, 23:25:

Hope these are clear, but this is essentially my 80 yellow IDE connector going from the top HDD (E) into IDE 1 and shows the adapter.

Thanks for the pictures. 4 things are going against you there--

1. Mixing the 40 pin cable and the 80 pin cable is not a good idea. The 80 pin cable will tell your device that faster than 33Mhz is OK, but the 40 pin portion of the cable is unlikely to be able to handle that.
2. The max length of IDE cables is supposed to be <= 18" (45cm). If you go over 18", the chance of signal problems increases.
3. When you use an extension cable or an adapter in the IDE cabling, it creates additional signal reflections, and the chance of signal problems increases.
4. When IDE cables are touching power cables, it creates additional signal challenges, and the chance of signal problems increases.

So the moral of this story:

1. If you use a single 80 conductor cable that is <= 18" and keep it away from the power cables, it will probably work fine.
2. With your current situation, maybe it works if you can force PIO 3, UDMA 0 or MWDMA 1. I'm not surprised that UDMA2 or faster has issues.

Awesome. Thanks for the clarification. I’m learning new stuff all the time. Especially regarding retro PCs. 🙂 Okay. So my suspicions were right regarding the 80 IDE connector with an adapter. But what would the solution be? As I say, the IDE 1 and 2 buses are 40 pin exactly. The 80 connectors are 39. The ideal situation would be to hook it to IDE 3 or 4, but there’s the possibility it’s acting up with them. 😕 I’m sure I had the drive hooked to IDE 4 last time. With an 80 connector. The Highpoint Technology BIOS screen doesn’t really give any options to configure them.

I’m not sure if those options are available in the BIOS - PIO 3, UDMA 0 or MWDMA - I’ll have to check the images in another thread, or wait until I’m back home. I don’t recall seeing them though.

Might be tricky keeping the cables away from each other as things are pretty packed in there. 😅

DustyShinigami wrote on 2025-09-10, 22:11:

But what would the solution be? As I say, the IDE 1 and 2 buses are 40 pin exactly. The 80 connectors are 39. The ideal situation would be to hook it to IDE 3 or 4, but there’s the possibility it’s acting up with them. 😕 I’m sure I had the drive hooked to IDE 4 last time. With an 80 connector. The Highpoint Technology BIOS screen doesn’t really give any options to configure them.

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20.

Don't use an IDE cable extender. While some people has success with them in the 286 to early 486 era when IDE ran at <= 8 Mhz, once you get into the PCI age where IDE runs faster, you will often see data corruption when you use them because the IDE cable extenders are just too far out of spec to be counted on to operate reliably. So don't use an IDE cable extender.

Use a single 80 conductor cable that is <= 18 inches long and your problems will go away.

douglar wrote on 2025-09-11, 03:07:

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20. […]
Show full quote

DustyShinigami wrote on 2025-09-10, 22:11:

But what would the solution be? As I say, the IDE 1 and 2 buses are 40 pin exactly. The 80 connectors are 39. The ideal situation would be to hook it to IDE 3 or 4, but there’s the possibility it’s acting up with them. 😕 I’m sure I had the drive hooked to IDE 4 last time. With an 80 connector. The Highpoint Technology BIOS screen doesn’t really give any options to configure them.

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20.

Don't use an IDE cable extender. While some people has success with them in the 286 to early 486 era when IDE ran at <= 8 Mhz, once you get into the PCI age where IDE runs faster, you will often see data corruption when you use them because the IDE cable extenders are just too far out of spec to be counted on to operate reliably. So don't use an IDE cable extender.

Use a single 80 conductor cable that is <= 18 inches long and your problems will go away.

Gotcha. No extension. I only fret about 39/40 pin thing because IDE 1 and 2 have 40 pins and the 80 cables have 39 holes with the 40th filled in. So… how would it connect properly? One of the pins wouldn’t be able to go in, right?

DustyShinigami wrote on 2025-09-11, 12:28:

douglar wrote on 2025-09-11, 03:07:

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20. […]
Show full quote

DustyShinigami wrote on 2025-09-10, 22:11:

But what would the solution be? As I say, the IDE 1 and 2 buses are 40 pin exactly. The 80 connectors are 39. The ideal situation would be to hook it to IDE 3 or 4, but there’s the possibility it’s acting up with them. 😕 I’m sure I had the drive hooked to IDE 4 last time. With an 80 connector. The Highpoint Technology BIOS screen doesn’t really give any options to configure them.

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20.

Don't use an IDE cable extender. While some people has success with them in the 286 to early 486 era when IDE ran at <= 8 Mhz, once you get into the PCI age where IDE runs faster, you will often see data corruption when you use them because the IDE cable extenders are just too far out of spec to be counted on to operate reliably. So don't use an IDE cable extender.

Use a single 80 conductor cable that is <= 18 inches long and your problems will go away.

Gotcha. No extension. I only fret about 39/40 pin thing because IDE 1 and 2 have 40 pins and the 80 cables have 39 holes with the 40th filled in. So… how would it connect properly? One of the pins wouldn’t be able to go in, right?

just clear the blocked pinhole - you can drill it out or use something hot (say a small nail) to melt the plastic...I've done both successfully in the past

PC Hoarder Patrol wrote on 2025-09-11, 13:55:

DustyShinigami wrote on 2025-09-11, 12:28:

douglar wrote on 2025-09-11, 03:07:

Stop worrying about the 39/40 thing. That's not an issue. IDE doesn't use pin 20.

Don't use an IDE cable extender. While some people has success with them in the 286 to early 486 era when IDE ran at <= 8 Mhz, once you get into the PCI age where IDE runs faster, you will often see data corruption when you use them because the IDE cable extenders are just too far out of spec to be counted on to operate reliably. So don't use an IDE cable extender.

Use a single 80 conductor cable that is <= 18 inches long and your problems will go away.

Gotcha. No extension. I only fret about 39/40 pin thing because IDE 1 and 2 have 40 pins and the 80 cables have 39 holes with the 40th filled in. So… how would it connect properly? One of the pins wouldn’t be able to go in, right?

just clear the blocked pinhole - you can drill it out or use something hot (say a small nail) to melt the plastic...I've done both successfully in the past

Huh. I take it there’s still a wire/conductor underneath? I just wouldn’t want to completely damage it. 😬 I don’t really have anything that would work, apart from the nozzle of the hot glue gun, but I suspect that would be too thick to go through.

A heated sewing pin held with some pliers works well in my experience.

That particular pin is usually N/C.

It even gets clipped (or needs clipped) on CF->IDE adapters.

DustyShinigami wrote on 2025-09-11, 16:28:

Huh. I take it there’s still a wire/conductor underneath? I just wouldn’t want to completely damage it. 😬 I don’t really have anything that would work, apart from the nozzle of the hot glue gun, but I suspect that would be too thick to go through.

Doesn't matter. The pin isn't used. You can cut off pin 20 from your other cable and it won't make a difference.

douglar wrote on 2025-09-11, 16:31:

DustyShinigami wrote on 2025-09-11, 16:28:

Huh. I take it there’s still a wire/conductor underneath? I just wouldn’t want to completely damage it. 😬 I don’t really have anything that would work, apart from the nozzle of the hot glue gun, but I suspect that would be too thick to go through.

Doesn't matter. The pin isn't used. You can cut off pin 20 from your other cable and it won't make a difference.

Okay, cool. I’ll have to see if I can find something, or a suitable method, for making a hole in it. Thanks. 🙂

DustyShinigami wrote on 2025-09-11, 16:43:

douglar wrote on 2025-09-11, 16:31:

DustyShinigami wrote on 2025-09-11, 16:28:

Huh. I take it there’s still a wire/conductor underneath? I just wouldn’t want to completely damage it. 😬 I don’t really have anything that would work, apart from the nozzle of the hot glue gun, but I suspect that would be too thick to go through.

Doesn't matter. The pin isn't used. You can cut off pin 20 from your other cable and it won't make a difference.

Okay, cool. I’ll have to see if I can find something, or a suitable method, for making a hole in it. Thanks. 🙂

Back in the day when stuff was new and free cables seemed endless, I would take plyers and yank the unnecessary pin. Just make sure you pluck the correct one.

douglar wrote on 2025-09-11, 19:13:

DustyShinigami wrote on 2025-09-11, 16:43:

douglar wrote on 2025-09-11, 16:31:

Doesn't matter. The pin isn't used. You can cut off pin 20 from your other cable and it won't make a difference.

Okay, cool. I’ll have to see if I can find something, or a suitable method, for making a hole in it. Thanks. 🙂

Back in the day when stuff was new and free cables seemed endless, I would take plyers and yank the unnecessary pin. Just make sure you pluck the correct one.

I’ll play it safe and not remove any. 😅 I’ll just make a hole so the entire thing can be fully connected securely.