18 Jan
2001
18 Jan
'01
7 p.m.
Goldarg <goldarg(a)dub.net> wrote:
I recently got a AT&T SCSI drive from a friend and
wanted to see if I could
Model: 97536DA
Option: STD
Ser No: 3001a97212
Handle With Care
From the model and serial number, I'd guess that
this drive was
made by Hewlett-Packard in the first week of 1990. Those look
like HP model and serial numbers.
http://www.tame.com/hp/hpinfo.htm suggests that a 97536DA is a
330MB ESDI disk drive, OEM version.
If you think it's SCSI, I wouldn't be surprised that it is
high-voltage differential SCSI; HP did that sort of thing.
-Frank McConnell
19 Jan
19 Jan
9:19 a.m.
Goldarg <goldarg(a)dub.net> wrote:
They were pretty common in both single ended and differential.
I'll bet it's differential SCSI.
--Bill
--
bpechter(a)monmouth.com | FreeBSD since 1.0.2, Linux since 0.99.10
Brainbench MVP | Unix Sys Admin since Sys V/BSD 4.2
Unix Sys.Admin. | Windows System Administration: "Magical Misery Tour"
I recently got a AT&T SCSI drive from a
friend and wanted to see if I could
Model: 97536DA
Option: STD
Ser No: 3001a97212
Handle With Care
From the model and serial number, I'd guess
that this drive was
made by Hewlett-Packard in the first week of 1990. Those look
like HP model and serial numbers.
http://www.tame.com/hp/hpinfo.htm suggests that a 97536DA is a
330MB ESDI disk drive, OEM version.
If you think it's SCSI, I wouldn't be surprised that it is
high-voltage differential SCSI; HP did that sort of thing.
-Frank McConnell 
2:48 p.m.
Bill Pechter skrev:
They were pretty common in both single ended and
differential.
I'll bet it's differential SCSI.
And now for the 10 000$ question: what /is/ differential SCSI? In what way
does it differ?
--
En ligne avec Thor 2.6.
Fernsehen ist das einzige Schlafmittel, das mit den Augen genommen wird.
--- Vittorio de Sica
4:23 p.m.
And now for the 10 000$ question: what /is/
differential SCSI? In what
way does it differ?
It may be a $10K question, but you get a free answer, and hopefully worth
every penny. :-)
Differential SCSI is electrically completely incompatible with "normal"
(single-ended) SCSI. Differential is now called High-Voltage
Differential, or HVD, to distinguish it from the new Low-Voltage
Differential (LVD).
Single-ended signalling means that each signal pair has one signal that
is at TTL level (under 0.8V for logic 0, above 2.0V for logic 1), and a
signal return at ground level. Note that the active signal never carries
a negative voltage with respect to the return. (In reality there may
be short undershoots due to "ringing"; one of the functions of termination
is to minimize this.)
Differential signalling actively drives both signals of a pair. For a logic
0, one signal (called "true") is driven low, and the other
("complement")
is driven high. For a logic 1, the reverse is done. Note that this means
that the true signal voltage can be either positive or negative with
respect to the complement signal.
Differential signalling is preferred because it is more reliable,
especially when used with long cables. Many high-end servers and
workstations (e.g., IBM, HP, Sun) used HVD for this reason.
Unfortunately, HVD uses exactly the same connectors as single-ended SCSI,
though it will not interoperate unless you use a converter (generally
fairly expensive).
The newer LVD operates on the same principle but at lower voltage levels.
LVD signalling is not compatible with either HVD or single-ended. However
(before everyone jumps all over me), LVD devices will detect that they
are plugged into a single-ended SCSI bus, and switch to single-ended
signalling. Thus you can mix LVD and single-ended SCSI devices on one
bus, but you do not get the benefits of LVD signalling when you do so.
20 Jan
20 Jan
1:45 a.m.
Eric Smith skrev:
> And now for the 10 000$ question: what /is/
differential SCSI? In what
> way does it differ?
It may be a $10K question, but you get a free answer,
and hopefully worth
every penny. :-)
Differential SCSI is electrically completely
incompatible with "normal"
(single-ended) SCSI. Differential is now called High-Voltage
Differential, or HVD, to distinguish it from the new Low-Voltage
Differential (LVD).
Single-ended signalling means that each signal pair has
one signal that
is at TTL level (under 0.8V for logic 0, above 2.0V for logic 1), and a
signal return at ground level. Note that the active signal never carries
a negative voltage with respect to the return. (In reality there may
be short undershoots due to "ringing"; one of the functions of termination
is to minimize this.)
Differential signalling actively drives both signals of
a pair. For a logic
0, one signal (called "true") is driven low, and the other
("complement")
is driven high. For a logic 1, the reverse is done. Note that this means
that the true signal voltage can be either positive or negative with
respect to the complement signal.
Differential signalling is preferred because it is more
reliable,
especially when used with long cables. Many high-end servers and
workstations (e.g., IBM, HP, Sun) used HVD for this reason.
Unfortunately, HVD uses exactly the same connectors as
single-ended SCSI,
though it will not interoperate unless you use a converter (generally
fairly expensive).
The newer LVD operates on the same principle but at
lower voltage levels.
LVD signalling is not compatible with either HVD or single-ended. However
(before everyone jumps all over me), LVD devices will detect that they
are plugged into a single-ended SCSI bus, and switch to single-ended
signalling. Thus you can mix LVD and single-ended SCSI devices on one
bus, but you do not get the benefits of LVD signalling when you do so.
Thank you for the exhaustive explanation. But does this mean that differential
SCSI doesn't have any ground pins, instead using these for the complement?
--
En ligne avec Thor 2.6.
Vi m?ste vara r?dda om varandra
- det ?r det enda reciproka pronomen vi har.
2:09 p.m.
does this mean that
differential SCSI doesn't have any ground pins, instead using these for
the complement?
There are some ground lines, but not nearly as many as in single-ended
SCSI. "Narrow" SCSI has 9 data lines (8+parity) and 9 control lines, so
there are 18 differential pairs.
Another interesting feature of HVD (high voltage differential) SCSI is
that there are no "active terminators". The terminations are just single
resistors (110 ohm? I forget.) between the signals of a pair. No
termination voltage source needed.
19 Jan
19 Jan
4:37 p.m.
At 02:48 20-01-2001 +0100, Iggy Drougge wrote:
And now for the 10 000$ question: what /is/
differential SCSI? In what way
does it differ?
Differential is merely a type of electrical interface. Its use is not
limited to SCSI devices: Certain types of high-speed serial ports, such as
V.35, are also differential.
Here's the details. Single-ended SCSI interfaces have one signal line, and
one matching ground return line, for each SCSI signal (REQ, ACK, D0-D7,
etc.) A signal line is considered to be active or 'asserted' when the line
goes to a logic high (TTL logic, +5V usually) in reference to the ground line.
With differential, you have two active signal lines per signal. When said
signal is asserted, one line will be pulled to a given positive voltage
(often +12) while the other is pulled to the opposite negative voltage (-12
in this context).
A differential interface only recognizes that a signal has gone active
when there is a significant difference in potential between the two lines.
This gives a diff interface much greater noise immunity than that of
single-ended setups, and it also allows much greater cable length than you
can have in SE (25 meters vs. 9, as I recall... someone check me on that,
please?)
Critical point: You CANNOT mix diff and SE SCSI devices on the same bus
without an interface converter on the mismatched device. Without said
converter, one of two things will happen.
1). At best, built-in protective features in the interface drivers will
detect the presence of a differential device (the DIFFSENSE line), and will
shut down the entire bus.
2). If, for some reason, said features fail to function as designed,
you'll be reaching for a fire extinguisher pretty quick.
Recommended reading: http://www.scsifaq.org
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Bruce Lane, Owner and head honcho, Blue Feather Technologies
http://www.bluefeathertech.com // E-mail: kyrrin(a)bluefeathertech.com
Amateur Radio: WD6EOS since Dec. '77 (Extra class as of June-2K)
"I'll get a life when someone demonstrates to me that it would be
superior to what I have now..." (Gym Z. Quirk, aka Taki Kogoma).
21 Jan
21 Jan
5:11 a.m.
New subject: IBM 399*
Was in the local recyclers and noticed 4 of these monsters
sitting on a palette. They said they had more. They are
about the size of a photocopier paper box, dark grey finned
cases with a IBM 399* ( the asterix was on the label) 40 kg
label on one end. Are these guts for a 3390? Nicely
shock mounted on big rubber mounts.
8584
days inactive
8586
days old
7 comments
6 participants
participants (6)
-
eric＠brouhaha.com -
fmc＠reanimators.org -
KenzieM＠sympatico.ca -
kyrrin＠bluefeathertech.com -
optimus＠canit.se -
pechter＠pechter.dyndns.org