On Sunday 15 September 2002 22:56, Jason Thomas wrote:
>On Sun, Sep 15, 2002 at 10:47:41PM -0400, Gene Heskett wrote:
>> On Sunday 15 September 2002 17:44, Jason Thomas wrote:
>> >Okay here are the details.
>> >
>> >
>> >Controller:
>> >00:09.0 SCSI storage controller: Adaptec AHA-2940U2/W
>> >
>> >The Tape Drive:
>> >Its a HP dat40x6i Autoloader. C5713A and connects directly to
>> > the 68 pin cable.
>> >
>> >The hard drives are seagate barracuda's and as far as I know
>> > have no built in termination.
>>
>> Ahh, but they do. Its an "active" termination when its enabled,
>> which is via jumper placement on the drive,
but no huge resistor packs like you may be used to seeing.
>>You'll have to
>> slide the drives out of the rails to see the jumper tables
>> printed on the top and the jumper locations are sometimes on the
>> front edge of the drive, and sometimes on the bottom, very
>> occasionally on the rear adjacent to the header socket.
>>
>> Is the tape the last device on the cable, and connected to the
>> last connector?
>
>controller ----- 50 pin cable ---- terminator of some sort.
>
> 68pin cable
>
> hdd
>
> tape
>
> hdd
>
> hdd
>
> active term
>
>cable is about 1M long.
I'm not familiar with that controller, some have a seperate 50 pin
buss for the rear panel, and some combine them but terminate the
upper 18 lines right beside the 50 pin hidens connector. I'd
assume that such is the case since I think you can check the state
of the terms with the ctrl-a option at bootup. I assume you have
done that, and that they are enabled or on. In which case you may
dispense with the external 50 pin "terminator of some sort" as that
would be overkill.
However, with the tape in the middle of an otherwise working bunch
of HD's, it becoming a bit of a stretch to totally blame the
terminations.
One last element maybe, (can you tell I'm grasping at straws here?)
have you run something like memcheck86 on that machine in recent
history?
Replaced that cable with a fresh one?
Subbed a different active term on the end of the cable?
Tried the tape on a different connector?
Unless I was there, with a decent scope to look that cable over
with, I'm fresh out of (remotely controllable) ideas. As thats a
decent bandwidth there, I'm not sure I'd trust a scope with less
than 500mhz of bandwidth (and decent probes rated for that
bandwidth, basicly one needs at least 10x the expected bandwidth,
100x would be nice but bring little red wagons loads of large bills
for that), and my own is only 100 mhz, dual trace, computerized.
And thats not enough to fully display echo's in a 1 meter cable. I
can see them, but cannot make definitive measurements. It's a < $3k
Hitachi.
Bed-thirty I think...
--
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz 512M
99.15% setiathome rank, not too shabby for a WV hillbilly
|
