IDE drives -- only one works at a time!

esuhl

(Yes I've checked the drive jumpers and IDE cable!)

This has had me baffled -- I can only put it down to some kind of motherboard failure, but I thought I'd run it past you guys in case anyone's seen anything similar.

I have a PC with four SATA drives, a PCIe SSD, an IDE HDD, and an IDE optical drive. There is only one IDE channel.

I recently swapped the IDE HDD out for a spare one, so I could install Windows 10 and snag the free upgrade licence, before replacing the original IDE HDD.

Since replacing the drive, only the "master" device on the IDE channel is recognised by the BIOS.

Originally, the drive jumpers were set to "cable select", with the optical drive as master (on the end of the IDE cable). If I disconnect either device, the other one is recognised fine in the BIOS -- so I know the drives both work.

I tried setting the jumpers specifically to master/slave and then swapping them over (so the optical drive was the slave), but in any case, only the master device is recognised.

I replaced the IDE cable with another one, and have exactly the same problem.

I already have the latest available BIOS version, and the IDE detection timeouts in the BIOS have been set to the maximum (35 seconds). I tried setting the timeout to zero (as I was unsure if this meant "wait for zero seconds" or "wait indefinitely"), but that didn't help.

The PCIe SSD is (strangely) labelled as an IDE drive in the BIOS and boot menu, so I tried removing that to see if the two "real" IDE drives would be recognised, but still no luck.

I can't think what else could be wrong, other than a motherboard fault. But it seems strange that the IDE master would work, but not the slave. Everything was working fine until I swapped the original drive back.

I don't suppose anyone has seen this before or has any suggestions?

bluesnake

you get slightly different ide layouts. Both have 40 pin connectors.

The first version (very, very old) used master/slave drive selected by a jumper on the drive. The "cable" was a common straight through one and any connector could be used. This had a 40 conductor ribbon cable.

The newer versions (very old) was a slightly different setup. The drives were set to cable select. The "cable" had either two or three connectors. These connectors had a specific designation and hence were coloured differently. This often had a 80 conductor ribbon cable.

bluesnake

you get slightly different ide layouts. Both have 40 pin connectors.

The first version (very, very old) used master/slave drive selected by a jumper on the drive. The "cable" was a common straight through one and any connector could be used. This had a 40 conductor ribbon cable.

The newer versions (very old) was a slightly different setup. The drives were set to cable select. The "cable" had either two or three connectors. These connectors had a specific designation and hence were coloured differently. This often had a 80 conductor ribbon cable.


Some drive models were not compatible with other drives on the same channel, but this was fairly infrequent, but did happen.

Personally I wild use ide for the CD only, and convert drive to sata to overcome this issue, but by the time you buy bits, it might be cheaper and more reliable just buying one big drive

Heedtheadvice

Odd that your ide combination worked before swapping but not when reinstated.
Did you have to change anything when you put in the temp Ide HDD? ...and possibly not reinstated? You mention they were both set to cable select but you don't mention if you have tried that config on the original cable but maybe have done. Perhaps that should be the first step details as above?

I presume that you have not loaded win10 but wonder if during the process it has changed anything.
What are you using to report the status of the connected drive recognition, thinking is this a windows thing?

. Do you have any 'test software ' for disk drives? Such as provided by the manufacturers of the drives? Some of these bypass windows running from their own system. That might take one variable out of the equation. Seagate and Western Digital and probably others supply them for free.

esuhl

bluesnake wrote: »

you get slightly different ide layouts. Both have 40 pin connectors.

The first version (very, very old) used master/slave drive selected by a jumper on the drive. The "cable" was a common straight through one and any connector could be used. This had a 40 conductor ribbon cable.

The newer versions (very old) was a slightly different setup. The drives were set to cable select. The "cable" had either two or three connectors. These connectors had a specific designation and hence were coloured differently. This often had a 80 conductor ribbon cable.

Some drive models were not compatible with other drives on the same channel, but this was fairly infrequent, but did happen.

I've been using the PC with the two IDE drives for around six years, so I imagine everything's compatible! :-)

The cable is one of the newer types with 80-wire cable and a blue end that plugs into the motherboard. I got rid of the ribbon-style cables ages ago. The ones I have look more like this:



I thought I only had four SATA ports, but I've actually got six, with one not being used, so I suppose I could get an IDE to SATA adaptor -- I just don't understand what could have happened to the IDE channel. :-/ It was fine before...

Heedtheadvice wrote: »

Odd that your ide combination worked before swapping but not when reinstated.
Did you have to change anything when you put in the temp Ide HDD? ...and possibly not reinstated? You mention they were both set to cable select but you don't mention if you have tried that config on the original cable but maybe have done. Perhaps that should be the first step details as above?

Yup -- I didn't change anything. Just unplugged the slave (set to cable-select), plugged in the spare drive (also set to cable-select), installed Win10, then put the old drive back in place. The optical drive remained on the end of the cable throughout.

Heedtheadvice wrote: »

I presume that you have not loaded win10 but wonder if during the process it has changed anything.

I installed Windows 10 successfully on the IDE drive. Win10 recognised the optical drive (and all my other drives) too. I played with it for a day or so, switching between Windows 10, Win7 and Linux (via BIOS and boot loader settings), then went back to the old drive.

Essentially, nothing has changed.

Heedtheadvice wrote: »

What are you using to report the status of the connected drive recognition, thinking is this a windows thing?

I'm using the BIOS POST. The IDE controller displays its own message listing the detected drives.

Heedtheadvice wrote: »

Do you have any 'test software ' for disk drives? Such as provided by the manufacturers of the drives? Some of these bypass windows running from their own system. That might take one variable out of the equation. Seagate and Western Digital and probably others supply them for free.

Alas that won't work since the drives aren't (both simultaneously) detected in the BIOS. No subsequently loaded OS detects them. The drives work fine individually, using the same (or a different) cable -- no issues or errors at all.

I'm stumped...

prowla

Hmm, it seems like you've tried the options (swapping cables, etc.)

What happens if you unplug all other drives and just keep the two IDE ones?

It's not possible that the master/slave/1/2 jumpers are wrong (or faulty), is it?

Can you set any to slave?

System

Hi

Yes, cable select is with a 'special cable'; think back to the old floppy cable that had a twist in several conductors.

http://www.unixwiz.net/techtips/ide-cable-select.html
http://www.pcguide.com/ref/hdd/if/ide/confCS-c.html

So the other thing is eyesight and orientation. I have old IDE drives where the label is a mirror image of the jumper pin layouts, and when they are faded/dirty my poor old eyes struggle to get my head around it.

spud17

Just a thought, when all else fails, try resetting the BIOS to defaults.

bluesnake

I recently swapped the IDE HDD out for a spare one, so I could install Windows 10 and snag the free upgrade licence, before replacing the original IDE HDD.

Simples answer: is to put the other drive back and if it starts to work again it is drive compatability issue. If it still does not work, but once did would suspect cable.

The compatibility issue was with the implementation if the ide interface of the different manufacturers built in controllers and not the mboard controller.

Questions 15, 16, 17
http://www.dewassoc.com/kbase/hard_drives/ide_eide_qa.htm#Q15

prowla

bluesnake wrote: »

Simples answer: is to put the other drive back and if it starts to work again it is drive compatability issue. If it still does not work, but once did would suspect cable.

The compatibility issue was with the implementation if the ide interface of the different manufacturers built in controllers and not the mboard controller.

Questions 15, 16, 17
http://www.dewassoc.com/kbase/hard_drives/ide_eide_qa.htm#Q15

The OP is saying that it did work with the drives in question, then they swapped one, and then they put the original ones back and it is now having issues with them.

esuhl

prowla wrote: »

It's not possible that the master/slave/1/2 jumpers are wrong (or faulty), is it?

Can you set any to slave?

I was thinking that -- a sudden jumper failure sounds unlikely, but it's about the only thing I haven't checked.

Neither drive seems to work as a slave now, either set to slave or "cable select".

johndough wrote: »

So the other thing is eyesight and orientation. I have old IDE drives where the label is a mirror image of the jumper pin layouts, and when they are faded/dirty my poor old eyes struggle to get my head around it.

The drives had been working fine for years, and I hadn't needed to change the jumper selections. The labels on both drives are quite clear and I've double, triple and quadruple checked them. :-/

bluesnake wrote: »

Simples answer: is to put the other drive back and if it starts to work again it is drive compatability issue. If it still does not work, but once did would suspect cable.

I'll see if I can find the Win10 drive again... I just shoved it back in my cupboard of computer bits!

It can't be a compatibility issue as it's been working like this for years. And it's unlikely to be the cable as I swapped it for another one and get the exact same issue.

bluesnake wrote: »

The compatibility issue was with the implementation if the ide interface of the different manufacturers built in controllers and not the mboard controller.

Questions 15, 16, 17
http://www.dewassoc.com/kbase/hard_drives/ide_eide_qa.htm#Q15

Interesting reading. I don't think the drives are so old that they came out before the IDE standard was fully ratified! (How old would they have to be?!)

As I said, the drives worked fine and I hadn't changed jumper positions before I started troubleshooting, but I had forgotten that there used to be different selections for "master" and "master with slave present" -- so maybe that's been preventing the slave from being detected during my subsequent tests, although it doesn't explain what suddenly stopped it working in the first place.

I had a few problems re-seating the IDE connectors initially, so I wonder if I might have powered up the PC with the drives not-quite-connected, and whether that might have caused a problem with the IDE controller that hasn't been rectified now the drives are properly plugged in...?

spud17 wrote: »

Just a thought, when all else fails, try resetting the BIOS to defaults.

Yeah... I might have to try that.

I've often noticed issues after hardware changes, which always seem to be rectified by just going into the BIOS and saving the settings (without changing them). I did this as a force of habit, but maybe resetting everything is the way to go.

I booted up once yesterday and the BIOS POST message from the IDE controller (which usually lists drive IDs along with a text description) listed the master drive as normal, but displayed the slave ID without a description (which I've never seen before). It only happened once. So maybe this is a CMOS setting problem.

Thanks for all the tips and suggestions, everyone! :T

I'll put the hammer away for now, and have another look later!