25 Nov
2012
25 Nov
'12
3:23 p.m.
On 11/24/2012 02:00 PM, Chris Elmquist wrote:
On Saturday (11/24/2012 at 02:06AM -0700), Eric Smith
wrote:
This is clearly using a tactical nuke to clear an ant farm. The original
TU-58 uses a very bare 8085 with only 256 bytes or ram and 2K of Rom
running at a clock speed that is 5/6ths the nominal speed of that CPU.
This should give a clear indication that very little processing power is
needed.
As to what a TU58 is, it is only two drives of 256K bytes each, with
serial interface.
The communications is serial to 38Kbaud and the protocol is very specific
but fairly simple.
For people with PDP11s this thing without tapes (solid state memory flavor)
is adequate to boot an 11 and even do useful thing s and without the slow
down (seek and rewind) of mechanical tapes it's fast enough to be very
useful.
For VAX11/750 and 730 users thsi is the microcode boot device and also
what is used to run diags. So it's important to have a tape or its analog.
Allison
I wrote:
I saved an image
of the ROM while I had this unit apart but I see you
guys are already way ahead.
If the raw tapes more readily available, I think it would make sense to
pursue the ability to format them. But since it seems even raw tapes
are approaching unobtainium in any quantity, maybe it is time to revisit
the solid-state emulation. Perhaps a job for a BeagleBone or Raspberry
Pi these days...
Chris It is remotely possible that there's an
undocumented "format"
command in the protocol. However, I've heard multiple people
claim that special firmware was required.
I spent some time reverse-engineering
the firmware. There is only
one undocumented opcode, decimal 10, and I haven't yet figured out
what it does, but it definitely doesn't format a tape.
I think I understand the code that writes a block, but of course
there's no code to write an address field. Writing formatter
firmware shouldn't be too hard, but it would be best to have a known
good tape to examine to find out the details of the gap lengths.