HP-1000 HP-IB / GP-IB IEEE-488.2, etc

[Tim Riker]

Thanx for the info!

J. David Bryan wrote:
> On 7 Apr 2008 at 16:48, Tim Riker wrote:
>> I understand that HP-7920 and HP-7925 drives were AMIGO flavor GP-IB
>> drives.
> 
> No, these drives used the MAC command set.

Got a pointer to docs on that?

>> I'm not quite sure where the GP-IP buss comes in the picture.
> 
> There were effectively three different interface systems employed by the 
> 79xx series of drives.
> 
> The 7905/06/20/25 drive systems debuted the MAC (Multi-Access Controller) 
> command set.  This used a 13037 controller box, one to eight drives, and 
> one to eight CPU interfaces.  There were master and slave versions of each 
> drive (e.g., 7925M, 7925S, etc.) -- the essential difference being that a 
> master drive came with the 13037 controller and the slave did not; the 
> drives otherwise were identical.  Interface to the HP 1000 was via the 
> 13175 card for the first computer and 13178 card for the rest (the only 
> differences were cabling and termination resistors on the 13175 card).

HP-1000 w/13175/13178 <cable-a> 13075 <cable-b> 7905/06/20/25 yes?

What are cable-a and cable-b? dual 50 pin card edge on both ends of both
cables?

> Later, an HP-IB option was added via the 12745 card.  This card was 
> inserted in the 13037 box and provided an HP-IB connector at the rear of 
> the chassis.  Connection to the HP 1000 was via the 12821 disc interface 
> card, rather than the 13175 card.  With the HP-IB option, the multi-CPU 
> capability was lost, although up to eight drives could still be connected 
> to the single CPU.  The MAC command set was retained, although it was now 
> transferred via HP-IB.  Note that the addition of the 12745A card did not 
> change the command set to CS/80.

So now we have:

HP-1000 w/12821 <cable-c> 13075 w/12745 <cable-d> 7905/06/20/25

where cable-c is a 50-pin to gp-ib? and cable-d is as cable-b above?

> Finally, to provide lower-cost installations, the "H" series of ICD 
> (Integrated Controller Disc) products was introduced (e.g., 7925H).  These 
> dispensed with the 13037 controller box and added a controller card within 
> the drive itself, presenting an HP-IB connector on the rear of the drive 
> chassis.  Up to four such drives could be connected together, and they used 
> the same 12821 interface for connection to a single HP 1000.  The MAC 
> command set was still employed, although three of the commands dealing with 
> multiple CPUs and error correction (the 13037 provided ECC; the ICD 
> controllers did not) were deleted.

HP-1000 w/12821 <cable e> 7925H

so I presume cable-e is 50 pin card edge to GP-IB?

You say the drives could be "connected together". Is this just a normal
GP-IB bus? or are there still a unit number here?

>> There is normally a contoller (13037-60028) interface...
> 
> That's the Device Controller PCA number, one of three cards within the 
> 13037 chassis (or four, if the 12745 HP-IB option were added).

ah!

>> ...between the drives and the host computer interface (02640-90042?)
> 
> I don't recognize that number at all, although 02640-xxxx should be a 264x 
> terminal part number.

I forget which pdf I read that number in. I think it was one of the
13037 documents. Can the 26xx terminals also speak GP-IB?

>> How does the shared storage through a 13037 controller work? are they
>> all just IDs on GB-IB? 
> 
> The 13037+12745 controller appeared as a single HP-IB device; the attached 
> drives were addressed via the "U" (unit number) field of the appropriate 
> MAC commands.  Each ICD controller had its own HP-IB address; the unit 
> number fields with the commands were not used.

I'm confused. See previous "connected together" comment.

Thanx for the reply! Things are getting clearer.

I picked up a National Instruments GPIB-SCSI converter. That could add
whole new monkey wrenches to the picture. :) The docs say it'll drive in
one direction or the other. ie: a machine with a scsi bus can talk to
gpib devices through it, or a machine with a gpib bus can see scsi
devices. It appears that it's one or the other, not both.
-- 
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
¿ǝʇɐɔıʇǝu pooƃ ǝɹnʇɐuƃıs ɹnoʎ uı 8-ɟʇn ƃuıʌɐɥ sı