I have two identical HP 36GB SCSI drives (68-pin connector) that I got 
from an auction site. When I brought them home and hooked them up to a 
Linux box, they both spun up and tested OK. I could mount them, copy 
files between them, etc. When I turned the computer on the next day (or 
maybe it was two days later), however, I noticed that only one was seen 
and spun up at start-up.

Taking a closer look, the other appears to be dead. Specifically, the 
system sees and spins up one drive, but not the other - regardless of 
whether I connect both at once or one at a time. It is not a termination 
or bus number collision problem. A couple days ago the controller 
firmware saw the drive and I could use it, and now it doesn't and I can't.

I noticed that some numbskull set a jumper across the pins for 
connecting a disk activity LED. Could this have caused the drive to die 
a premature death?

Any other suggestions on what I could look at to further diagnose the 
problem?

Thanks,
Nathan

smithfarm wrote:
> 
> [...]
> I noticed that some numbskull set a jumper across the pins for
> connecting a disk activity LED. Could this have caused the drive to die
> a premature death?

Not very likely but not impossible. The LED has several volts forward
voltage, if there is no real current regulation for the LED on the
controller but only a simple resistor, the current over the jumper can
be twice the value that was intended for the LED. But a good design
should foresee this condition and tolerate it.

> Any other suggestions on what I could look at to further diagnose the
> problem?

You wrote that the disk has an 68pin connector, this means there is also
a 4pin connector for power supply. Because of the high mechanical forces
on the 4 power supply pins I have seen disks where the (surface mount)
solder connection between pin and controller board was broken. The
cracks in the solder can be nearly invisible, so take a second look. It
helps to plug in the power connector and bend it in both directions
while looking at the solder joints.


Micha

smithfarm <presnypreklad@gmail.com> wrote:
> I have two identical HP 36GB SCSI drives (68-pin connector) that I got 
> from an auction site. When I brought them home and hooked them up to a 
> Linux box, they both spun up and tested OK. I could mount them, copy 
> files between them, etc. When I turned the computer on the next day (or 
> maybe it was two days later), however, I noticed that only one was seen 
> and spun up at start-up.
> 
> Taking a closer look, the other appears to be dead. Specifically, the 
> system sees and spins up one drive, but not the other - regardless of 
> whether I connect both at once or one at a time. It is not a termination 
> or bus number collision problem. A couple days ago the controller 
> firmware saw the drive and I could use it, and now it doesn't and I can't.
> 
> I noticed that some numbskull set a jumper across the pins for 
> connecting a disk activity LED. Could this have caused the drive to die 
> a premature death?
> 
> Any other suggestions on what I could look at to further diagnose the 
> problem?

Do you expect to fix the drive after diagnosing why it failed?

like any 10 year old hard drive, it belongs in the trash.