VOGONS


First post, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie

Hi Vogons-Retro Community,

I have a rather odd problem, regarding a WD Caviar WD400 Hard Drive I've used as a primary Hard Drive in my Win98-System on a Pentium 3 Motherboard (Intel 440 ZX, Aopen AX6BC-EZ), both connected to a Promise Ultra66 or directly to the onboard Intel IDE-Controller, jumpered as Master.

It just isn't detected anymore. BIOS detection hangs, no hard drive is found. It happened without any typical noises made by dying hard drives.

I've tried several cables (old and new, all known to be working). PSU has been upgraded for a Voodoo 5, it worked before with a weaker PSU so I doubt it is a power issue either. Tests with less power hungry 3dfx cards don't change my hard drives misbehavior.

A test on another, more modern motherboard (Asus A8n-SLI) confirms it is definitely not a size barrier issue (the Aopen AX6BC-EZ detects HDD up to 128 GB just fine).

The Hard Drive was extracted and built into an IcyBox IDE-USB Case for wiping and preparing it for another project. Both Windows 10 and MX Linux (Debian based) detection over USB is fine - the drive showed 3 Partitions (MSDOS file table, 1 Primary, 1 extended Partition with to logical partitions inside). I wiped it both under WinX and Linux with a new msdos filesystem table. After it wouldn't be detected even with a new file system table, I did a "diskpart clean" to the drive over USB. This also had no effect on both retro PCs. Detection over USB is still working though.

In one of my tests the BIOS didn't detect the hard drive, but a windows xp setup (SP3, known to be working) run for testing purposes detected the hard drive and offered to setup a C: Partition and format it in NTFS Format. This worked, but somewhat slowly. After copying the installation files from CD-R and asking to reboot, the system crashed BADLY, it wouldn't even show any video output.

This gave me the idea, that the drive had physical issues. So I've screwed off the PCB an looked if there were any corroded contacts - it was all like new! No broken pins, no corrosion.

I'm am out of ideas. Any idea what I could do - other than swap out the hard drive 😁

Too lazy for a signature. Imagine one 😉

Reply 2 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie

Yes, it does. I sees my Samsung 160 GB HDD up to 136 GB without any issue. The same goes for my promise controller. I guess that means, I'm out of luck 🙁
Thanks anyway 😀

Too lazy for a signature. Imagine one 😉

Reply 4 of 23, by RetroGamer4Ever

User metadata
Rank Member
Rank
Member

Indeed, the jumpers may be the issue. I had to learn that the hard way, back when there was no Internet to consult with. As soon as I could move to SATA, I never looked back at IDE, though I still keep an IDE card and some cables in my toolbox, for whatever purpose they could serve in the future.

Reply 6 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie

Hi, my drive doesn't have a jumper setting for single drive. For single drive it's supposed to be configured as master.

However I'll definitely try cable select and 2 drives on one channel - I missed that when checking the drive. Thanks for the suggestions!

Too lazy for a signature. Imagine one 😉

Reply 7 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie
RetroGamer4Ever wrote on 2022-01-28, 16:28:

Indeed, the jumpers may be the issue. I had to learn that the hard way, back when there was no Internet to consult with. As soon as I could move to SATA, I never looked back at IDE, though I still keep an IDE card and some cables in my toolbox, for whatever purpose they could serve in the future.

SATA is indeed a so much more convenient way of hooking up drives. I really like it a lot, especially since it's capable of Hot Plug, AHCI and so on, but for a retro build it's supposed to be at least as much the way it was back in the day as I can justify my time and money for it.

If it weren't for some headache because of some configuration quirks, some fun in playing with old parts would go missing for me 😁

Too lazy for a signature. Imagine one 😉

Reply 8 of 23, by the3dfxdude

User metadata
Rank Member
Rank
Member
Kampfkoloss wrote on 2022-01-28, 17:44:

Hi, my drive doesn't have a jumper setting for single drive. For single drive it's supposed to be configured as master.

However I'll definitely try cable select and 2 drives on one channel - I missed that when checking the drive. Thanks for the suggestions!

It's likely a WD drive has a jumper setting for a single drive. It's been that way for years. What is connected to your IDE cable? If it's the only drive, remove the jumper and place it on the storage position. They usually don't print the instructions on the drive label.
https://eshop.macsales.com/Tech/manuals/idehd … ttings/WD72.pdf

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

Reply 9 of 23, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie
the3dfxdude wrote on 2022-01-29, 15:39:

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

I try both, especially with USB adapters I often have more luck with the Cable Select option.

Reply 10 of 23, by hyoenmadan

User metadata
Rank Member
Rank
Member

Your drive "service area" cylinder got corrupted. You may give on it, as the only reliable way to fix this is rebuilding the SA with some specialized HDD analysis software like the ACELabs PC3000 (yes, you have to rebuild it, a simple copy from a donor will destroy the drive), and then reuploading the rebuilt cylinder dump to its corresponding place. And yeah, you may hope the SA doesn't contain bad sectors, or then all this still will not work.

Reply 11 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie
the3dfxdude wrote on 2022-01-29, 15:39:

It's likely a WD drive has a jumper setting for a single drive. It's been that way for years. What is connected to your IDE cable? If it's the only drive, remove the jumper and place it on the storage position. They usually don't print the instructions on the drive label.
https://eshop.macsales.com/Tech/manuals/idehd … ttings/WD72.pdf

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

My 98-PC has a single HDD. I've now tried the single drive jumper setting I've found in the PDF you provided... And it worked! I didn't know, they didn't print their complete jumper diagram on their hard drives - what a mess o0
The only settings displayed on my HDD (Caviar WD400) are Master/Single ::|:: Slave :::|: and Cable Select ::::| (: = free pins, |= jumper attached).

What really messes with my mind is: it worked before, with this HDD jumpered as Master in the very same PC.

Too lazy for a signature. Imagine one 😉

Reply 12 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie
Disruptor wrote on 2022-01-29, 17:28:
the3dfxdude wrote on 2022-01-29, 15:39:

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

I try both, especially with USB adapters I often have more luck with the Cable Select option.

I also tried cable select with the hard drive connected to my promise pci controller with a 80 wire-IDE cable and it worked fine, too with the jumper settings provided in the PDF. I can't judge if cable select is more reliable, I haven't tried it in any of my PCs containing IDE HDDs up until now 😁

Too lazy for a signature. Imagine one 😉

Reply 13 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie
hyoenmadan wrote on 2022-01-30, 00:59:

Your drive "service area" cylinder got corrupted. You may give on it, as the only reliable way to fix this is rebuilding the SA with some specialized HDD analysis software like the ACELabs PC3000 (yes, you have to rebuild it, a simple copy from a donor will destroy the drive), and then reuploading the rebuilt cylinder dump to its corresponding place. And yeah, you may hope the SA doesn't contain bad sectors, or then all this still will not work.

Is it still possible this area is broken, when the drive is being detected again be the mainboard and the promise controller's bios? If so, I think it wouldn't be reliable as a system disk anymore.

Too lazy for a signature. Imagine one 😉

Reply 14 of 23, by hyoenmadan

User metadata
Rank Member
Rank
Member
Kampfkoloss wrote on 2022-01-30, 19:18:

Is it still possible this area is broken, when the drive is being detected again be the mainboard and the promise controller's bios? If so, I think it wouldn't be reliable as a system disk anymore.

Yes, is possible if the SA has weak bits...
There is also a possibility on the head amplifiers being on their last.

Reply 15 of 23, by the3dfxdude

User metadata
Rank Member
Rank
Member
Kampfkoloss wrote on 2022-01-30, 19:12:
the3dfxdude wrote on 2022-01-29, 15:39:

It's likely a WD drive has a jumper setting for a single drive. It's been that way for years. What is connected to your IDE cable? If it's the only drive, remove the jumper and place it on the storage position. They usually don't print the instructions on the drive label.
https://eshop.macsales.com/Tech/manuals/idehd … ttings/WD72.pdf

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

My 98-PC has a single HDD. I've now tried the single drive jumper setting I've found in the PDF you provided... And it worked! I didn't know, they didn't print their complete jumper diagram on their hard drives - what a mess o0
The only settings displayed on my HDD (Caviar WD400) are Master/Single ::|:: Slave :::|: and Cable Select ::::| (: = free pins, |= jumper attached).

Good. Now does the drive work reliably? Western Digital Drives have always been a bit vague on the single drive setting, unless you had kept the original instructions. They are not the same as other drives.

Kampfkoloss wrote on 2022-01-30, 19:12:

What really messes with my mind is: it worked before, with this HDD jumpered as Master in the very same PC.

You may have been using cable select, or using cable select somewhere else. It sometimes work, with the right cable/controller. I never really bothered fully understanding that arrangement, since I started with IDE in the earliest days and it was usually just set master or slave. I don't see why cable select was a good idea. At best it does work, but sometimes not in the order you want for it, so you have to move drives around rather than just set jumpers.

I still have a very early western digital IDE drive here. I'm not sure if it has the single/dual thing there, it's been a long time since I used it, but it still works.

Kampfkoloss wrote on 2022-01-30, 19:16:

I also tried cable select with the hard drive connected to my promise pci controller with a 80 wire-IDE cable and it worked fine, too with the jumper settings provided in the PDF. I can't judge if cable select is more reliable, I haven't tried it in any of my PCs containing IDE HDDs up until now 😁

Yup. Well I probably wouldn't bother testing CS in other configurations now that you know how to set a WD drive. But anyhow, I think the question now is if the drive is actually on the last legs or not. The last WD400 Caviar I tried a few weeks ago was dead, and I've seen a few WD drives dead or dying from that era. Not sure about reliability of WD from this time period (2001-2004?) at this point. When's the manufacture date?

Reply 16 of 23, by rasz_pl

User metadata
Rank Oldbie
Rank
Oldbie
Kampfkoloss wrote on 2022-01-30, 19:18:
hyoenmadan wrote on 2022-01-30, 00:59:

Your drive "service area" cylinder got corrupted. You may give on it, as the only reliable way to fix this is rebuilding the SA with some specialized HDD analysis software like the ACELabs PC3000 (yes, you have to rebuild it, a simple copy from a donor will destroy the drive), and then reuploading the rebuilt cylinder dump to its corresponding place. And yeah, you may hope the SA doesn't contain bad sectors, or then all this still will not work.

Is it still possible this area is broken, when the drive is being detected again be the mainboard and the promise controller's bios? If so, I think it wouldn't be reliable as a system disk anymore.

No, hyoenmadan post would only make sense if drive stopped working altogether, or was working intermittently. Working in USB enclosure is a clear sign drive is ok, but there is problem with detection. Detecting IDE drives is a mess because for historical reasons. You often couldnt even use more than one drive from different manufacturer on same cable back in early nineties

http://www.os2museum.com/wp/the-dual-drive-ide-hell/
>To behave correctly, Device 0 has to act differently depending on whether Device 1 is present; if Device 1 is present, Device 0 has to let it respond to register accesses directed to it, but if there is no Device 1, then Device 0 has to respond to simulate a two-drive controller with no second drive attached.

USB enclosure doesnt care and doesnt even try accessing second device, therefore bypasses all potential problems.

hyoenmadan wrote on 2022-01-30, 21:23:

Yes, is possible if the SA has weak bits...
There is also a possibility on the head amplifiers being on their last.

A lot of things are possible. This drive could be possessed by ghosts, developed AI, or contain contents of whole Library of Congress. What makes you think its storage mechanism is defective?

Reply 17 of 23, by PC Hoarder Patrol

User metadata
Rank Oldbie
Rank
Oldbie
Kampfkoloss wrote on 2022-01-30, 19:18:

Is it still possible this area is broken, when the drive is being detected again be the mainboard and the promise controller's bios? If so, I think it wouldn't be reliable as a system disk anymore.

I'd suggest you try running WDs Data Lifeguard Diagnostics / Tools from DOS to check on the drives health - https://web.archive.org/web/20050413004116/ht … downloadxml.asp

Reply 18 of 23, by hyoenmadan

User metadata
Rank Member
Rank
Member
rasz_pl wrote on 2022-01-31, 05:40:

A lot of things are possible. This drive could be possessed by ghosts, developed AI, or contain contents of whole Library of Congress. What makes you think its storage mechanism is defective?

Experience with the WD4xx and WD8xx drive series. Back then was noted the erratic behavior described by OP, and this detail about being recognized only on certain USB enclosures (back then were trascend usb ide bridges the best ones for the work) and apparently working well with them for some time, until the platter data wasn't longer accessible (on BIOS you would see a 10MB or 0MB drive, with the drive internal "nickname" as drive identifier string, and a random generated serial number). Autopsy on these drives revealed one physical cause (dead head amplifier chips), and multiple software causes (SA data corrupted (common with certain nvidia, ALi, and the dreaded VIA KT133 chipset), or SA weak bits (which in turn was a sign something was wrong with the head amplifier chip).

Reply 19 of 23, by Kampfkoloss

User metadata
Rank Newbie
Rank
Newbie
the3dfxdude wrote on 2022-01-31, 01:56:
Good. Now does the drive work reliably? Western Digital Drives have always been a bit vague on the single drive setting, unless […]
Show full quote
Kampfkoloss wrote on 2022-01-30, 19:12:
the3dfxdude wrote on 2022-01-29, 15:39:

It's likely a WD drive has a jumper setting for a single drive. It's been that way for years. What is connected to your IDE cable? If it's the only drive, remove the jumper and place it on the storage position. They usually don't print the instructions on the drive label.
https://eshop.macsales.com/Tech/manuals/idehd … ttings/WD72.pdf

I find cable select unreliable. I always jumper the drive specific to the configuration, plus respect the Western Digital way of jumpering.

My 98-PC has a single HDD. I've now tried the single drive jumper setting I've found in the PDF you provided... And it worked! I didn't know, they didn't print their complete jumper diagram on their hard drives - what a mess o0
The only settings displayed on my HDD (Caviar WD400) are Master/Single ::|:: Slave :::|: and Cable Select ::::| (: = free pins, |= jumper attached).

Good. Now does the drive work reliably? Western Digital Drives have always been a bit vague on the single drive setting, unless you had kept the original instructions. They are not the same as other drives.

Kampfkoloss wrote on 2022-01-30, 19:12:

What really messes with my mind is: it worked before, with this HDD jumpered as Master in the very same PC.

You may have been using cable select, or using cable select somewhere else. It sometimes work, with the right cable/controller. I never really bothered fully understanding that arrangement, since I started with IDE in the earliest days and it was usually just set master or slave. I don't see why cable select was a good idea. At best it does work, but sometimes not in the order you want for it, so you have to move drives around rather than just set jumpers.

I still have a very early western digital IDE drive here. I'm not sure if it has the single/dual thing there, it's been a long time since I used it, but it still works.

Kampfkoloss wrote on 2022-01-30, 19:16:

I also tried cable select with the hard drive connected to my promise pci controller with a 80 wire-IDE cable and it worked fine, too with the jumper settings provided in the PDF. I can't judge if cable select is more reliable, I haven't tried it in any of my PCs containing IDE HDDs up until now 😁

Yup. Well I probably wouldn't bother testing CS in other configurations now that you know how to set a WD drive. But anyhow, I think the question now is if the drive is actually on the last legs or not. The last WD400 Caviar I tried a few weeks ago was dead, and I've seen a few WD drives dead or dying from that era. Not sure about reliability of WD from this time period (2001-2004?) at this point. When's the manufacture date?

Hey, sorry for the late reply. The drive's from 1 Nov 2003. I've got the drive used a few years ago without its original documentation and packaging and it has served in many iterations of this project reliably.

It's now set up in neutral position / single drive as provided in the PDF by 3dfxdude. I think CS is a bit of a hassle, because -as far as I understand- all other drives in the system have to support CS AND have to be jumpered accordingly. Without original documentation it's hit or miss.

Back to my HDD:
After shortly working when connected to mainboard and promise ultra66 pci (not at the same time, but one after another) without any other cards except for voodoo and sb live (sb0100), the whole system froze again after adding my other cards to the system (intel 1000MT-NIC, Via 6212L USB2 Controller).

I think the drive is either on its last legs OR it's a major resource conflict because of too many cards.

Too lazy for a signature. Imagine one 😉