Er... are we talking 40-pin desktop/3.5' IDE or 44-pin notebook/2.5' here?
If 40-pin it should be easy to test using a desktop PSU to see where the fault lies...
If 44-pin then perhaps knowing the model / seeing a picture might lead to some clues.
I know some 1.8' drives used 3.3v...

I use my USB adapter for any more modern drives -- let's say 20GB+. But, they're definitely not fool-proof. It has not been able to reliably detect most (maybe all) of my older drives -- like 2GB and below, particularly. It did not work with my ATAPI ZIP drive. It works with most CD-ROMs OK, but failed miserably on my Creative Hex-Speed CD-ROM.

So here's what I do:

I have a Pentium II running a modern Gentoo Linux (took FOREVER to compile the kernel), and installed a removable drive bay. When I get a new IDE HDD, or just want to test one before using it again, I'll drop it in there and run badblocks on it. I also use smartmontools to check the counters before and after a round of surface testing to see how healthy the drive appears to be. The box is also networked, so I can use dd to copy disk images to and from my NAS.

I still have to figure out how to test optical drives. I'm sure there's a good Linux utility for doing measured reads from disc. I need something that can help me determine whether the laser(s) is/are still good. I will also have to figure out something convenient for attaching them, preferably without having to pop the side panel.

SirNickity wrote:

I still have to figure out how to test optical drives. I'm sure there's a good Linux utility for doing measured reads from disc. I need something that can help me determine whether the laser(s) is/are still good.

It's not really intended for that purpose, but dvdisaster's Scan option might be useful there.

BushLin wrote:

If 40-pin it should be easy to test using a desktop PSU to see where the fault lies...

I know, but I don't have a reliable (=100% sure that it works) PSU to check that. Also, I'm reducing my collection, so I'm looking for less bulky things at home. The disk that gave me problems has been working for about two months, so obviously I had some kind of problem at my testing equpment.

BushLin wrote:

I know some 1.8' drives used 3.3v...

IDE or SATA? My machines use 2.5 and 3.5 inch HDD drives , and most of them only provides +5 and +12v. I was wondering it fhere are new disks that can't be fitted on that machines becase of power requirements.

SirNickity wrote:

I have a Pentium II running a modern Gentoo Linux (took FOREVER to compile the kernel), and installed a removable drive bay. When I get a new IDE HDD, or just want to test one before using it again, I'll drop it in there and run badblocks on it. I also use smartmontools to check the counters before and after a round of surface testing to see how healthy the drive appears to be. The box is also networked, so I can use dd to copy disk images to and from my NAS.

As I said, I'm trying to use less bulky hardware. I actually use that adaptor using smatrmontools to check counters and/or run SMART tests (if I need my main computer, I connect it to a laptop or Raspberry Pi). If I need to make a more thorough test, I use Victoria (lastest versions can run tests on USB disks). I've tried whdd (using System Rescue CD), but can't run a reliable surface test on USB bus (seriously, I miss MHDD).

SirNickity wrote:

I still have to figure out how to test optical drives. I'm sure there's a good Linux utility for doing measured reads from disc. I need something that can help me determine whether the laser(s) is/are still good. I will also have to figure out something convenient for attaching them, preferably without having to pop the side panel.

I guess you'll need some kind of special hardware to do that. You want to measure laser power, and the drives only detect if their laser is working or not.

A possible test would be using a special disk that features different oppacity in different zones and some known pattern on it. Imagine a disk that reflects 100% light at sectors 1 to 1000, 95% between 1001 and 2000 and so on. If you try to read that disk, at some point the laser won't produce enough reflected light to read accurately and will fail. So knowing where it failed and a threshold value (i.e.: a good laser should read at least x percent in reflectivity) you will know if your laser is good, unreliable or is going to fail soon.

I don't know if that kind of disks or tests exists, so...

How about a docking station?

For instance: Docking Station.

Zup wrote:

A few days ago, I had to check an old IDE hard disk before installing it. I connected to my chinese USB to IDE/SATA adapter, and the thing didn't spin up. The disk was detected, the hard disk ID was consistent (so the board worked), but the size was 0 bytes and I couldn't read or write anything to it and I couldn't hear the motors revving up. I tested with jumpers on MASTER, MASTER (SINGLE) and CS positions, and limiting it to 32 Gb but I couldn't make it work.

Did you, like, connect a desktop 3.5" drive to the USB to IDE? If so, did you connect the external power? 3.5" drives require 12V, and USB does not provide that. This can explain why the motors did not spin up.

eL_PuSHeR wrote:

How about a docking station?

For instance: Docking Station.

I've got one of those. It works great with SATA disks (again... will it provide +3.3v on SATA power?), but it doesn't support IDE disks. I found only one dock that supports IDE, this one from Startech. The problem is that it is very expensive, costing more than 80€ (a SATA dock station only costs about 30€).

dr_st wrote:

Did you, like, connect a desktop 3.5" drive to the USB to IDE? If so, did you connect the external power? 3.5" drives require 12V, and USB does not provide that. This can explain why the motors did not spin up.

It was connected to the computer via USB, and to the cheap-cheap-cheap PSU that came with the cable. As I said, that PSU only had three wires on the molex connector (yellow, red and only one black wire) and it gave correct voltage when tested without being connected. I couldn't test it with the disk connected, so I suspect that either the hard disk drew enough power to overload the PSU and cause voltage to drop (it can be caused by a defective disk or a defective PSU) or it needed separate +5v and +12v lines (so it couldn't work because one of them had no ground connection).

Ah, the 3 wire PSU - some drives do NOT like it... get a molex fan adapter or molex splitter, and bridge the two blacks together.
Some drives obviously common the 0V and work, others don't, and will not run without seeing a 12V return.

I've had to use a bridge with the molex to SATA as well, as I've found SATA drives that are also inoperable without it.

Never found any SATA, not even SSD, that couldn't work on a molex to SATA

ssokolow wrote:

SirNickity wrote:

I still have to figure out how to test optical drives. I'm sure there's a good Linux utility for doing measured reads from disc. I need something that can help me determine whether the laser(s) is/are still good.

It's not really intended for that purpose, but dvdisaster's Scan option might be useful there.

Looks similar to Nero's Disc Speed util, which I do like for this purpose, but at least the version I have won't run on Win9x. So if this does, that's at least a partial win. Thanks! Would still like to get a Linux util too, though.

Zup wrote:

I guess you'll need some kind of special hardware to do that. You want to measure laser power, and the drives only detect if their laser is working or not.

No, nothing that special. I just want to know whether the drive is able to read a disc consistently. I have a couple drives that are not totally bad, but sometimes they'll just hang on a read, and Windows gives me that "insert the disc XXXX - if you didn't yank it out of there all sneaky-like, it may be damaged - sucks to be you" screen. I'm looking for a way to identify those drives so I can pull them apart and give them a good cleaning and re-lubing.

Zup wrote:

As I said, I'm trying to use less bulky hardware.

So, IIUC, you're really hoping to find a USB adapter that works with older or more finicky drives? Ah, I'm in the same boat then. FWIW, the most recent one I got (to support >2TB drives and USB 3.0) came with a wall-wart PSU with both Molex and SATA power connectors molded onto the cable. Handy, and seems to be reasonably solid. I haven't tried using that USB adapter with older pre Ultra-DMA disks yet, because it is temporarily occupied in my NAS to give me the last SATA interface I needed for RAID. I need to replace one more older, smaller disk, then I can resize the array to fit the 4 onboard SATA ports on that motherboard. Hopefully it works with old drives, but I'm not expecting much. Doesn't seem to be a priority these days.

SirNickity wrote:

ssokolow wrote:

SirNickity wrote:

I still have to figure out how to test optical drives. I'm sure there's a good Linux utility for doing measured reads from disc. I need something that can help me determine whether the laser(s) is/are still good.

It's not really intended for that purpose, but dvdisaster's Scan option might be useful there.

Looks similar to Nero's Disc Speed util, which I do like for this purpose, but at least the version I have won't run on Win9x. So if this does, that's at least a partial win. Thanks! Would still like to get a Linux util too, though.

Linux is dvdisaster's primary supported platform and it's what I use it on.

OH. I totally missed that. I saw all the Windows screen shots and put it on the to-do list to try on my pre-XP boxen. Thanks, that's fantastic!

SirNickity wrote:

OH. I totally missed that. I saw all the Windows screen shots and put it on the to-do list to try on my pre-XP boxen. Thanks, that's fantastic!

Yeah. It'll probably be in your distro's package repo too.

https://repology.org/project/dvdisaster/versions

So, at last, I could hook a proper PSU and test the hard disk (and another disk from the same model).

The disk was not incompatible with my adaptor, it was dying.

The HDD (a Hitachi Deskstar HDS728080PLAT20) did work when I pulled it from a machine, but when I tested it at home it didn't spin up at all. My computer detected it, but the size parameter was 0 or -1 (it showed 4294 million blocks) and Linux complained that the big was too disk. The other disk I tested was detected correctly using the same PSU and USB adaptor.

So I guess my hardware was running fine (but I remember it failed with a disk drive smaller than 1 Gb that worked connected in a 486... maybe the disk was not totally ATA?).

Thanks for all the help/suggestions.

I think your assumption is correct. That's why I use a removable drive bay in one of my older PCs. Several old disks I tried just don't work on my USB to IDE adapter, even when they work fully on an actual IDE interface.