4 Feb
2014
4 Feb
'14
9:59 p.m.
On Tue, Feb 4, 2014 at 10:39 PM, David Gesswein <djg at pdp8online.com> wrote:
Slower than planned but I have now released the MFM
emulator code
and updated the MFM reading code. Needs more testing but it
seems to work.
http://www.pdp8online.com/mfm/mfm.shtml
Cool! Well done!
1) Gold edge fingers increase the board cost. Would
people prefer a cheaper
board with .1 headers like my wirewrap board that will need non standard
cables or proper edge connectors for the drive emulation?
How much extra? It's not that expensive to make 20-pin and 34-pin
cables that replace the originals, but it is handy to have a drop-in
board where the real drive goes since usually there are already
cables. Would it be worth considering optional snap-on or solder-on
paddle interfaces or would that just cost more in the end than it's
worth? At least they'd be small.
2) It might be possible to make the board emulate two
drives. The
system would have to be setup to use daisy-chained control cable since the
board will only have one control connector and two data connectors. The second
data cable would have to be .1 header. Is this enough interest to people
to be worth trying to do?
I can see how that might be handy on an RQDX3, but for me at least on
non-DEC systems, I have little need to try to max out what the
controller can do. If it weren't too expensive, I might use it, but
if it adds a double-digit percentage to the cost, I might not need it
that badly.
In terms of my personal systems, I'm only really likely to "need" an
MFM-type drive in a 3B1/PC7300, anything with a DEC interface
(DECmate, RQDXn, Professional series...) especially if this is cheaper
than a Qbus SCSI card, and maybe a CBM D90x0 hard drive unit. I have
MFM controllers for x86 machines but am not as interested in this for
that target - there are usually other routes for disks in PC clones.
I also have a small number of MFM controllers for Amigas, but again,
there are other ways of putting storage on those that don't require
emulating an analog interface.
-ethan