4 Feb
2014
4 Feb
'14
10:39 p.m.
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
Next to test with will be my VAXStation II assuming I can figure out how to
run it.
I'm also staring on the PCB design to replace my wirewrap board. A couple
of questions for people interested. Two people have said they are
building wirewrap/protoboards themselves.
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?
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?
5 Feb
5 Feb
12:59 a.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
7:27 a.m.
On Tue, 4 Feb 2014, David Gesswein 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.
Outstanding. I've been wanting something like this for quite a while.
The immediate problem I'm facing is how to replace the IMI drives in my
Corvus flat-cable systems. They are dropping left and right, but
something about the interface scheme prevents it from working with
anything other than the IMI units. Seagates will go through the motions
but don't verify, acting as if they have not actually been formatted.
--
7 Feb
7 Feb
8:44 a.m.
On Wed, Feb 05, 2014 at 07:27:49AM -0500, Steven Hirsch wrote:
The immediate problem I'm facing is how to replace the IMI drives in
my Corvus flat-cable systems. They are dropping left and right, but
something about the interface scheme prevents it from working with
anything other than the IMI units. Seagates will go through the
motions but don't verify, acting as if they have not actually been
formatted.
What drive model/interface does that use? I did a little searching and
some IMI units use a non ST506 interface so without changes my unit can't
deal with them. The IMI 77xx are neat looking drives.
9:47 a.m.
On Fri, 7 Feb 2014, David Gesswein wrote:
On Wed, Feb 05, 2014 at 07:27:49AM -0500, Steven
Hirsch wrote:
>
> The immediate problem I'm facing is how to replace the IMI drives in
> my Corvus flat-cable systems. They are dropping left and right, but
> something about the interface scheme prevents it from working with
> anything other than the IMI units. Seagates will go through the
> motions but don't verify, acting as if they have not actually been
> formatted.
What drive model/interface does that use? I did a
little searching and
some IMI units use a non ST506 interface so without changes my unit can't
deal with them. The IMI 77xx are neat looking drives.
Corvus used the IMI 5006, 5012 and 5021H depending on the storage
capacity. From everything I can see, these are (or should be) plain
vanilla MFM drives. However, when I substituted a known-good Seagate
drive with equivalent geometry it refused to actually work. As I
mentioned in my original post, it goes through the motions of formatting
(steps through tracks for a decent amount of time) but then immediately
fails verification. I'm guessing there's a timing descrepancy or some
behavioral difference behind this.
The drive controller is also an IMI product that is not based on a
dedicated HDC chip. It's uses a Z80 and a bunch of peripheral logic,
probably a proprietary encoding scheme.
The more I think about this, the more I think they've taken some pains to
ensure it will only work with IMI drives - maybe some sort of special
operational mode?
Steve
--
5 Feb
5 Feb
7:50 a.m.
On 2014-02-04 20:39, David Gesswein wrote:
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?
I think the proper edge connectors would be worth paying more for, for
certain values of "more".
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?
Marginally useful for an RQDX3 type of setup, but probably not worth
much
extra cost... I can't imagine too many uses for (for instance) multiple
RD-54's, given the use cases of your typical retrocomputing
enthusiasts,
but I could (of course) be wrong.
Congrats on your progress! Your design will help many, I'm sure.
--
/*
* John P. Willis
* Coherent Logic Development
* Solutions: Everywhere
*
* Office: 575.524.1034
* Mobile: 575.649.4318
*
* jpw at coherent-logic.com
* http://www.coherent-logic.com/
*
*/
6 Feb
6 Feb
7:57 a.m.
On Wed, Feb 05, 2014 at 05:50:55AM -0700, John P. Willis wrote:
Marginally useful for an RQDX3 type of setup, but probably not worth
much
extra cost... I can't imagine too many uses for (for instance) multiple
RD-54's, given the use cases of your typical retrocomputing
enthusiasts,
but I could (of course) be wrong.
My VAXstation II has a distrubition board with a separate control cable
to each drive after the RQDX3 so I was thinking this was a system that
my proposed multiple drive support wouldn't work in since it required a
single control cable. If the board just sends the same signals to each drive
and merges the lines back it might still work.
Most people seem to think not of that much interest. The cost would be
minor since the parts wouldn't need to be populated. So if it doesn't
take much time and low risk I'll put the hardware support on the board.
Software support for it will not be a high priority.
5 Feb
5 Feb
9:40 a.m.
From: David Gesswein, Tuesday, February 04, 2014 7:39 PM
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?
I get an added cost of less than $2 per board for the gold fingers (from the
quote engine at sunstone.com). At that price, I think it would be worth it to
be able to use the vintage cables folks presumably already have, rather than
new cables that would probably cost more than $2 to make.
Vince
10:33 a.m.
On Feb 4, 2014, at 10:39 PM, David Gesswein <djg at pdp8online.com> wrote:
...
1) Gold edge fingers increase the board cost.
I think that depends on the board supplier you use. In these days of the lead-free
religion and surface mount devices, a thin gold plating is a standard finish for boards,
and the shop I?ve been doing business with (PCB-pool) offers it as a no-cost option.
That?s probably not as thick of strong as the classic gold plated board fingers, but it
may well be good enough for the small number of insert/remove cycles likely with hobbyist
use.
paul
12:27 p.m.
> 1) Gold edge fingers increase the board cost.
1) In the days of bad connections on TRS80 floppy card edges, somebody was
making an IDC MALE 34 pin card edge, that was used along with an ordinary
34 pin female IDC card edge connector to make very short 34 pin card edge
to card edge extenders. If those still exist, then the board could have
dual row header, with a short cable.
2) In the early days of 3.5" floppy drives, when most computer floppy
cables were card edge, there were 34 pin dual row header to card edge
adapters. Most were a tiny circuit board, but there were a very few that
were a molded adapter.
5:18 p.m.
On 5 February 2014 15:33, Paul Koning <paulkoning at comcast.net> wrote:
On Feb 4, 2014, at 10:39 PM, David Gesswein <djg at pdp8online.com> wrote:
between my vintage machines, and only having a small number of emulated
disks means I might have to move the emulated disk(s) around even more, so
I think a hobbyist might actually end up with more insert/remove cycles
than is normal for a production machine.
Regards
Rob
...
1) Gold edge fingers increase the board cost.
I think that depends on the board supplier you use. In these days of the
lead-free religion and surface mount devices, a thin gold plating is a
standard finish for boards, and the shop I've been doing business with
(PCB-pool) offers it as a no-cost option.
That's probably not as thick of strong as the classic gold plated board
fingers, but it may well be good enough for the small number of
insert/remove cycles likely with hobbyist use.
paul
I don't know about you, but I find that I often have to move disks around
3818
days inactive
3820
days old
10 comments
8 participants
participants (8)
-
cisin＠xenosoft.com -
djg＠pdp8online.com -
ethan.dicks＠gmail.com -
jpw＠coherent-logic.com -
paulkoning＠comcast.net -
robert.jarratt＠ntlworld.com -
snhirsch＠gmail.com -
vrs＠msn.com