IDE drives -- only one works at a time!

esuhl

I'm sure everyone's lost interest by now, but I think the problem is related to a problem with the JMicron PATA controller in the BIOS.

I tried every combination of drives, jumper settings, and 3 different cables. In every case, whenever I use the original 60GB drive, the slave device failed to be detected (regardless of whether the 60GB drive is master or slave).

I've been running the PC for a few days with everything connected as it was originally (optical master, original 60GB drive as slave). Earlier, I modified an unrelated BIOS setting, booted up and, by an apparent miracle, both drives appeared! That only happened once, though. Every reboot since, only the master drive was detected. Argh!

So, I thought it would be worth replacing the CR2032 battery and resetting the BIOS settings to the default values (using a jumper on the motherboard).

On the next boot, the BIOS POST message listed both drives, but the slave was in red, flashing text. I rebooted and both drives were listed in their normal colour. Subsequent boots leave me with just the master drive again.

I can't work out why the behaviour isn't consistent, why it had been working fine for years and has now failed, and whether there's anything I can do to fix it. (I'm guessing probably not.)

It's the strangest, most annoying IT problem I've come across.

paddyrg

esuhl wrote: »

It's the strangest, most annoying IT problem I've come across.

...so far...there's still time!

Fightsback

Check the appropriate P/O for CS on the mobo, it could be dry jointed, worth a try. It's does happen.

http://pinouts.ru/HD/AtaInternal_pinout.shtml

esuhl

Fightsback wrote: »

Check the appropriate P/O for CS on the mobo, it could be dry jointed, worth a try. It's does happen.

http://pinouts.ru/HD/AtaInternal_pinout.shtml

Sorry... I'm not sure what you mean. I understand what a "pin out" is, and a dry joint, but what's CS and how can I verify the pin out -- using a multimeter?

(I hope I don't have to resolder anything -- I'll make one heck of a mess!)

The IDE interface on the motherboard is mounted at right-angles to the board, which seems a bit silly. Plugging and unplugging the cable is probably putting a strain on the soldered joints. :-/

---

Just one other thing... The drive contains raw data with no partitions or filesystems. Maybe the PATA controller (stupidly) requires a partition to detect a drive...?

When I first attached it, the drive did have partitions, so maybe it was detected and remembered by the BIOS even when I subsequently reformatted (unformatted?!) the drive. Maybe I need to put a tiny partition at the front of the drive to give it something to detect...?

Fightsback

esuhl wrote: »

Sorry... I'm not sure what you mean. I understand what a "pin out" is, and a dry joint, but what's CS and how can I verify the pin out -- using a multimeter?

CS - Cable select, fully read the link I posted all the way to the bottom

Just a thought, you did unplug the IEC PSU cable and wait a couple of mins before swapping IDE drives ?

Heedtheadvice

You really are getting into the more remote probabilities - but still possibilities - now and maybe the particular nuances of the BIOS and mobo. (The drive without partitions is recognised solely by the bios and so it should be otherwise how could you generate partitions in the first instance?!) I expect it is time you went on to a specialist forum to get specific advice about your system such as (similar issue) http://www.tomshardware.co.uk/forum/26510-32-hard-disk-detected-bios and post a question there withe details of the history, bios, mobo drives and versions. You get some really good help there.....or other similar forums!

Fightsback

Heedtheadvice wrote: »

http://www.tomshardware.co.uk/forum/26510-32-hard-disk-detected-bios

Hmmm a 404, that's helpful.

esuhl

Fightsback wrote: »

CS - Cable select, fully read the link I posted all the way to the bottom

Thanks. I did read all that -- I just wasn't quite sure what you were asking. I just don't understand how to check the pinout... surely all IDE interfaces are wired up the same? And how can I check for a dry joint? Sorry -- I'm fine with plugging stuff together, but this is possibly a little over my head.

Hang on... do you mean I should check the cable? I definitely have a suitable IDE cable with different coloured connectors (blue for the motherboard, grey for the slave, black for the master on the end). And it's definitely an 80-wire cable. I've tried three different cables with the same results, so I don't think it's a cable fault.

The PC had been running with the optical as master (as mentioned in that article) and (because I couldn't see why you'd set them to anything less than the maximum unless there was a problem), the BIOS timeouts for the IDE drives have always been set to the maximum of 35 seconds.

And I've tried the jumpers using master/slave instead of "cable select", and tried different jumpers, but the behaviour is pretty consistent..

Everything worked perfectly with exactly the same setup until I removed/replaced the 60GB drive. :-/

The only difference is that the first time the drive was connected, it contained partitions and filesystems (so maybe the BIOS detected and remembered it). Now it has raw (encrypted) data. I'm wondering if I need to wipe the drive and stick a partition on it for the (stupid, buggy) JMicron PATA controller to see it again...?

Fightsback wrote: »

Just a thought, you did unplug the IEC PSU cable and wait a couple of mins before swapping IDE drives ?

To all intents and purposes...

The PSU has a power switch. Before doing anything inside the PC, I switch the PSU off (so earth remains connected), then hit the power button repeatedly and wait for the motherboard LED to fade out before doing anything. I ground myself to the chassis/earth.

esuhl

Heedtheadvice wrote: »

You really are getting into the more remote probabilities - but still possibilities - now and maybe the particular nuances of the BIOS and mobo.

But... The BIOS has been fine for years running in this setup. All I did was briefly swap out a drive, then put the original back in.

I'll have a look at some more specialist forums, though. Cheers.

Fightsback

esuhl wrote: »

But... The BIOS has been fine for years running in this setup. All I did was briefly swap out a drive, then put the original back in.

I'll have a look at some more specialist forums, though. Cheers.

Check the the solder joints on the IDE cable plugs on your IDE drives as well.