20 Jan
2023
20 Jan
'23
4:13 p.m.
How do you know if a termimation is suitable? Is connection all you have to worry about? I
have an HP Ultra 320 drive, a 320/m compliant adapter (id jumpered to 2. Does a this need
to be 0 for a single drive setup?), the cable with an ultra 320m terminator (" LVD +
SE ACT NEG + HVD ISO " printed on it). Everything seems legit. I want to plug this
into 2 different serverboards, an Intel SCB2, dual PIII, dual ultra 160/lvd channels, and
an IBM xseries 350/Netfinity 6000 (8682 serverboard), quad PIII xeon slot 2 cpu's,
similar scsi capability.
Whaddaya think?
20 Jan
20 Jan
7:24 p.m.
I would read up to confirm the correct drive number so that variable is
confirmed.
Bill
On Fri, Jan 20, 2023, 4:13 PM Chris via cctalk <cctalk(a)classiccmp.org>
wrote:
How do you know if a termimation is suitable? Is
connection all you have
to worry about? I have an HP Ultra 320 drive, a 320/m compliant adapter (id
jumpered to 2. Does a this need to be 0 for a single drive setup?), the
cable with an ultra 320m terminator (" LVD + SE ACT NEG + HVD ISO " printed
on it). Everything seems legit. I want to plug this into 2 different
serverboards, an Intel SCB2, dual PIII, dual ultra 160/lvd channels, and an
IBM xseries 350/Netfinity 6000 (8682 serverboard), quad PIII xeon slot 2
cpu's, similar scsi capability.
Whaddaya think?
7:50 p.m.
Ok that. But I was told somewhere by someone that not all terminators are created equal.
I'd have to go look up the discussion.
On Friday, January 20, 2023, 07:24:50 PM EST, Bill Degnan via cctalk
<cctalk(a)classiccmp.org> wrote:
I would read up to confirm the correct drive number so that variable is
confirmed.
Bill
On Fri, Jan 20, 2023, 4:13 PM Chris via cctalk <cctalk(a)classiccmp.org>
wrote:
How do you know if a termimation is suitable? Is
connection all you have
to worry about? I have an HP Ultra 320 drive, a 320/m compliant adapter (id
jumpered to 2. Does a this need to be 0 for a single drive setup?), the
cable with an ultra 320m terminator (" LVD + SE ACT NEG + HVD ISO " printed
on it). Everything seems legit. I want to plug this into 2 different
serverboards, an Intel SCB2, dual PIII, dual ultra 160/lvd channels, and an
IBM xseries 350/Netfinity 6000 (8682 serverboard), quad PIII xeon slot 2
cpu's, similar scsi capability.
Whaddaya think?
9:14 p.m.
Indeed not not all SCSI terminators are created equal, there where three
types of signalling used Single Ended (SE) High Voltage Differential
(HVD) and Low Voltage Differential (LVD). LVD was created to squeeze a
bit more speed out of parallel SCSI. Any complaint LVD device needs to
be able to operate in SE mode as well as LVD and if they detect a SE
device on the bus all devices on the bus switch to SE mode, which is why
your terminator is marked LVD + SE. The other thing that LVD devices
are supposed to do is if they detect a HVD device on the bus the LVD
device isolates from the bus so that would be the "+ HVD ISO". SE
devices on the bus can be detected because one signal line is pulled to
ground by SE devices but not LVD, I don't recall which pin it is as
parallel SCSI is in what now seems the distant past. If your drive has
built in terminators and the terminator on you cable is not crimped on,
you could enable the termination on the drive and plug the drive at the
end of the cable in place of the terminator.
One thing to watch out for on your disk drive is many LVD drives have a
jumper to force SE mode you will want to remove that, although if force
to SE mode it should still work but will be slower, what does not work
is introducing a SE device into a bus that is already operating in LVD
mode, all the devices suddenly switching to SE mode generally causes the
using system to fall over.
You SCSI adapter card should always be set to address 7 as that is the
highest priority address and what you set the address on your drive to
will depend on. I have seen SCSI adapter that start scanning the bus at
0 and stop as soon as they do not get a response in which case you would
want your drive to be 0, most well designed adapter will scan all
addresses regardless, but if you plan to boot off the SCSI disk they may
have a limitation on the address the card will boot from. The most
important thing is that all device on the bus, including the adapter
have a unique address.
Paul.
On 2023-01-20 20:50, Chris via cctalk wrote:
Ok that. But I was told somewhere by someone that not all terminators are created equal.
I'd have to go look up the discussion.
On Friday, January 20, 2023, 07:24:50 PM EST, Bill Degnan via cctalk
<cctalk(a)classiccmp.org> wrote:
I would read up to confirm the correct drive number so that variable is
confirmed.
Bill
On Fri, Jan 20, 2023, 4:13 PM Chris via cctalk <cctalk(a)classiccmp.org>
wrote:
How do you know if a termimation is suitable? Is
connection all you have
to worry about? I have an HP Ultra 320 drive, a 320/m compliant adapter (id
jumpered to 2. Does a this need to be 0 for a single drive setup?), the
cable with an ultra 320m terminator (" LVD + SE ACT NEG + HVD ISO " printed
on it). Everything seems legit. I want to plug this into 2 different
serverboards, an Intel SCB2, dual PIII, dual ultra 160/lvd channels, and an
IBM xseries 350/Netfinity 6000 (8682 serverboard), quad PIII xeon slot 2
cpu's, similar scsi capability.
Whaddaya think?
21 Jan
21 Jan
6:58 a.m.
Paul Berger via cctalk wrote:
Indeed not not all SCSI terminators are created equal, there where three
types of signalling used Single Ended (SE) High Voltage Differential
(HVD) and Low Voltage Differential (LVD). LVD was created to squeeze a
bit more speed out of parallel SCSI.
That's not exactly correct. The reason for differential signalling was to be
able to use longer cables for a given speed. As speeds increased, the
useable length of an SE SCSI cable became prohibitively short and LVD
was devised to overcome this in a way that was relatively compatible with SE.
Also, just like what is now known as World War I was originally the Great War,
what we now known as HVD was originally known as just "differential SCSI".
Older equipment which is labelled "differential" is likely to be what we
now call HVD and connecting it to older SE only equipment may not work out well.
Regards,
Peter Coghlan.
551
days inactive
552
days old
4 comments
4 participants
participants (4)
-
Bill Degnan -
Paul Berger -
Peter Coghlan -
skogkatt007＠yahoo.com