Thank you. I see now why the older machine is working as it is.
I had put a Promise ATA100 controller board in it. Each hard drive,
C and D, is running off it's own Promise IDE channel.
Other (CD) drives run from the MB IDE33 controller.

I'd like to do the same with the new machine. I'd like to run my
C and D by themselves from the 2 built in MB IDE channels and
add a PCI controller board for other drives, CD, CDRW, etc.
The lights should then work independently. The present CD reader
and writer seem plenty fast on the IDE33 channels so I don't think
speed is critical. Would this work OK, and what type of controller
board would work with the least problems?  Thanks, Bob

Subject: [PCBUILD] Dual drive activity lights

> > I'm upgrading to an Intel D845PESV MB with WinXPPro.
> > It's in a Lian Li case with two Lian Li RH-321 hard drive mobile racks.
> > The top drawer. Whenever the computer is on,
> > both HD activity lights flash in unison all the time, from bootup until
> > shutdown. I have used this same setup on W98 machines for a long time and
>never saw
> > this. The D: drive always used to "sleep" until I Ghosted C: to the D: for
> > backup purposes. Anybody know why this is happening?
>
>I see it all the time, nothing wrong with it.  It happens when you have two
>devices on same cable and both mobile racks will flash the LED regardless
>which device is active, as long as there is activities going on on the
>channel, LED flashes.
>
> > Is this a false
> > indication?
>
>You can call it that way, but it is just what it was designed for.  All IDE
>racks I've seen do it (cheapo or expensive), I can't remember if SCSI racks
>have same problem or not, because I almost always setup SCSI in RAID or on
>its own (so can't tell).
>
> > Are both drives being written to identically?
>
>absolutely no.
>
> > Should they be on different IDE
> > controllers? Is this an XP or a hardware problem?
>
>As long as you keep the mobile racks, you will see it, non operating system
>or any other hardware is responsible for it.  I'm not sure why didn't you
>see it under win98, as it always happens.  If you keep the two racks on
>different channels, the dual flash will disappear, but you have to keep the
>two channels clean (only racks, no CDROM or any other IDE devices), as long
>as there is activities (eg. from CDROM), the rack LED will flash.
>
>Jun Qian