IBM channel-attached DASD emulation
tmfdmike at gmail.com
Sat Dec 26 08:03:09 CST 2015
On Dec 27, 2015 2:36 AM, "Noel Chiappa" <jnc at mercury.lcs.mit.edu> wrote:
> > From: Guy Sotomayor
> > It's typical of most of the vintage gear. Folks save the CPUs and
> > the peripherals.
> It's not entirely sloth and stupidity, though. Disk drives in particular
> (usually the biggest issue in this area) are complex precision machinery
> operate at very high speeds, etc, and working on them is a formidable job,
> and requires specialized parts which are, in general, no longer available.
> > I think the biggest problem is that there isn't a spec per-se on
> > Massbus. A lot of reverse engineering will be required to make it
> > properly in all cases.
> If it were done, though, that would be wonderful, especially for people
> PDP-11/70's; the UNIBUS on those machines is reputedly the slowest of any
> PDP-11, so having mass storage on the MASSBUS is really necessary for good
> performance. (Apologies for my -11 centrism in an IBM-focused thread...
No apologies needed.
Yea more: pdp-10s. Massbus is the *only* option for those. There are KS10
CPUs in particular - including mine - lying idle for want of Massbus
storage. I'm nearly there; I have a beautiful RM03 which should work - but
no bootable packs. I have a bunch of -10 RP06 packs - but no drives. I
don't have an operable Massbus tape.
> But finding the connectors (and probably the cables too) is going to be a
> cast-iron nightmare. Maybe we could settle on an alternative (the way I
> we should switch to pairs of dual cards with Berg/DuPont headers, with
> standard flat cables between them, to replace the now-unobtainable
> DEC showed this works, with the M9014/M9042..)
Certainly no need to slavishly reproduce the connectors and cables.
Emulation could plug in at the Berg level, direct to the Massbus
controller - or even at the board level, plugging straight into the
backplane (c.f. Guy's MEM-11) and replace the controller too.
More information about the cctalk