I have a CompuPro 816 System with a Q540 MFM 40 Meg HD that is getting
shaky. Sure would like to be able to use an IDE of which I have
several that would work. Maybe Someone will consider creating a MFM to
IDE Adapter Board that would insert between the MFM Cables and the IDE
Drive. I'm sure there is a lot of Older CP/M machines out there that
would benefit.
Bob in Wisconsin

"BrightFame"  wrote in message 
news:1fc72e56-6383-4ca9-80a6-9cd1dfc3cffe@8g2000prb.googlegroups.com...

I have a CompuPro 816 System with a Q540 MFM 40 Meg HD that is getting
shaky. Sure would like to be able to use an IDE of which I have
several that would work. Maybe Someone will consider creating a MFM to
IDE Adapter Board that would insert between the MFM Cables and the IDE
Drive. I'm sure there is a lot of Older CP/M machines out there that
would benefit.
Bob in Wisconsin

You might be able to get one here:

http://www.s100computers.com/My%20System%20Pages/IDE%20Board/My%20IDE%20Card.htm

Tom L

"Tom Lake" posted:
>
> You might be able to get [a S-100 hard disk controller?] here:
>
> http://www.s100computers.com/My%20System%20Pages/IDE%20Board/My%20IDE...

This is an IDE interface for S-100, not a MFM to IDE "converter" as
Bob (Bright Fame) requested. But it's the more practical solution
because it exists, and the magic "MFM to IDE" only existed as a
several hundred dollars commercial product. Search comp.os.cpm
archives to find the long history of discussing why "MFM to IDE" is
not simple.

But supporting IDE interfaces on Z80 systems has been done since IDE
became available; and discussed here in comp.os.cpm on almost an
annual basis. IDE is a pure digital interface and the IDE drive does
almost all the heavy lifting. Again, old discussions walk through
this.

John's project page above, fails to mention the decades-long history
of GIDE in the CP/M world. Versions of this small IDE interface by
Tilmann Reh of Germany have been available since the 1990's. I have a
Web page on the GIDE, because I first imported the GIDE from Europe
into the USA, in the 1990's; I choose to follow its progress, as does
Tilmann on his site.

http://www.retrotechnology.com/herbs_stuff/gide.html

Various IDE implementations have been done many times for Z80 and
other microcomputers. The Web page mentions some. It can be done with
a 8255 parallel interface chip, if your CPU is reasonably fast. Many
old S-100 cards have that chip and it's not hard to add one "from
scratch". But you can build an IDE interface with the same effort,
probably including the one at s100-computers.com.

And, recent discussion here on a microSD flash card controller with
SERIAL interface (not SPI) offers another hardware-simple and cheaper
solution which only requires a UART/serial port.

Herb Johnson
retrotechnology.com

On Thu, 24 Feb 2011 09:40:15 -0800 (PST), Herbert Johnson
<herbrjohnson@gmail.com> wrote:

>This is an IDE interface for S-100, not a MFM to IDE "converter" as
>Bob (Bright Fame) requested. But it's the more practical solution
>because it exists, and the magic "MFM to IDE" only existed as a
>several hundred dollars commercial product. Search comp.os.cpm
>archives to find the long history of discussing why "MFM to IDE" is
>not simple.
It would be much more straightforward to design a SASI/SCSI-to-IDE
interface, rather than capture the analogue high-speed differential
signalling at the MFM end.

I'm not saying it would be "simple", but translating the digital (and
inherently deterministic) interface would be an order of magnitude
less difficult. It would result in much faster (and probably more
reliable) and MUCH lower-powered data access.

It's surprising (to me, with my limited vocabulary of almost entirely
SASI-only and SCSI-only HDD Z-80 interfaces) that this option hasn't
been done to death already. Which means it must have been! :)

-PCP

On 02/24/2011 03:21 PM, PC Pete wrote:

> It's surprising (to me, with my limited vocabulary of almost entirely
> SASI-only and SCSI-only HDD Z-80 interfaces) that this option hasn't
> been done to death already. Which means it must have been! :)

These most definitely exist.  I have one in my Alpha UP2000+ system.  The box 
has an Adaptec 39160 SCSI controller and the converter lets me put a 
100-something GB IDE drive on line.

I've seen them for 50-pin SCSI as well.

Steve

On Thu, 24 Feb 2011 18:55:22 -0500, Steven Hirsch <snhirsch@gmail.com>
wrote:
>These most definitely exist.  I have one in my Alpha UP2000+ system.  The box 
>has an Adaptec 39160 SCSI controller and the converter lets me put a 
>100-something GB IDE drive on line.
>
>I've seen them for 50-pin SCSI as well.
Steve, do you remember who made your device? More to the point, do you
know if the manufacturer is still "alive"?

If there were such a beast designed for SASI, that would be a bonus.
If not, I've still got my SASI I/F schematics that might be worth
digging into.

I know a few Pulsar LBB system owners who would definitely be
interested in extending the life/reliability of their HDD systems!

If only I had another 12 or 13 hours a day, I'd jump in and design
something, no questions asked. I had to write a SASI drive exerciser
in assembler, and I still have the source (in printed format at
least). So I'm not too rusty. Ah well, if it isn't already a real
product, I'll add it to the list of stuff I need to design - alongside
the 8" to 5" FDD adapter, STD bus monitor, 2797 upgrade, etc!

If anyone else has seen anything similar to a SASI-to-IDE adapter
(especially if it was or is available in Australia), I'd be really
interested to hear from youse!

Cheers,
PCPete

On 02/25/2011 03:33 AM, PC Pete wrote:
> On Thu, 24 Feb 2011 18:55:22 -0500, Steven Hirsch<snhirsch@gmail.com>
> wrote:
>> These most definitely exist.  I have one in my Alpha UP2000+ system.  The box
>> has an Adaptec 39160 SCSI controller and the converter lets me put a
>> 100-something GB IDE drive on line.
>>
>> I've seen them for 50-pin SCSI as well.

> Steve, do you remember who made your device? More to the point, do you
> know if the manufacturer is still "alive"?

The unit I used was an ACARD AEC-7726H.  Judging from the product matrix on 
the box, you would want an ARS-2000FU or AEC-7720U.  The first is an entire 
5.25" frame + adapter, the second is a card with enhanced firmware to support 
optical drives (CD/DVD) in addition to hard disks.

I found quite a number of sources in a quick Google search.  Unfortunately, 
these devices have become (in my opinion) hideously expensive - perhaps 4-6x 
what I recall paying for my units.

Check out acard's web site.  All models appear current, albeit with usurious 
price tags.

Steve

Steven Hirsch  wrote:

> The [SCSI to IDE] unit I used was an ACARD AEC-7726H.

> I found quite a number of sources in a quick Google search. =A0Unfortunat=
ely,
> these devices have become (in my opinion) hideously expensive - perhaps 4=
-6x
> what I recall paying for my units.

I think I saw maybe two manufacturers, and several dealers of them.
Dealer prices were cheaper.

THere may have been a time when SCSI to IDE items were made by several
companies, in volume. So there was competition and volume to keep
prices down. Even so, current prices (I saw) of $150-$250 are not
awful. I recall MFM to (Something) converters were SEVERAL hundred
dollars in an earlier era. Steve, did you really pay under $50 for
your Acard converter, and if so when?

Herb Johnson
retrotechnology.com

On 02/25/2011 01:00 PM, Herbert Johnson wrote:
> Steven Hirsch  wrote:
>
>> The [SCSI to IDE] unit I used was an ACARD AEC-7726H.
>
>> I found quite a number of sources in a quick Google search.  Unfortunately,
>> these devices have become (in my opinion) hideously expensive - perhaps 4-6x
>> what I recall paying for my units.
>
> I think I saw maybe two manufacturers, and several dealers of them.
> Dealer prices were cheaper.
>
> THere may have been a time when SCSI to IDE items were made by several
> companies, in volume. So there was competition and volume to keep
> prices down. Even so, current prices (I saw) of $150-$250 are not
> awful. I recall MFM to (Something) converters were SEVERAL hundred
> dollars in an earlier era. Steve, did you really pay under $50 for
> your Acard converter, and if so when?

I definitely recall them being in the vicinity of $40-50 when purchased, which 
would have been about 6-7 years ago.

In <1dce94a4-4a76-4cc5-ab1f-3de1aca97381@y30g2000prf.googlegroups.com> Herbert Johnson <herbrjohnson@gmail.com> writes:

> [...]
> I recall MFM to (Something) converters
> [...]

   Do you mean ST-506 when you say "MFM"?

   Eddi ._._.
-- 
"Do whatever you want as long as you're not destroying somebody else's
 words." (Tim Skirvin)        -          e-mail: dk3uz AT arrl DOT net

On Feb 24, 3:21=A0pm, PC Pete <PCP...@audiography.com.au> wrote:
> On Thu, 24 Feb 2011 09:40:15 -0800 (PST), Herbert Johnson
>
> <herbrjohn...@gmail.com> wrote:
> >This is an IDE interface for S-100, not a MFM to IDE "converter" as
> >Bob (Bright Fame) requested. But it's the more practical solution
> >because it exists, and the magic "MFM to IDE" only existed as a
> >several hundred dollars commercial product. Search comp.os.cpm
> >archives to find the long history of discussing why "MFM to IDE" is
> >not simple.
>
> It would be much more straightforward to design a SASI/SCSI-to-IDE
> interface, rather than capture the analogue high-speed differential
> signalling at the MFM end.
>
> I'm not saying it would be "simple", but translating the digital (and
> inherently deterministic) interface would be an order of magnitude
> less difficult. It would result in much faster (and probably more
> reliable) and MUCH lower-powered data access.
>
> It's surprising (to me, with my limited vocabulary of almost entirely
> SASI-only and SCSI-only HDD Z-80 interfaces) that this option hasn't
> been done to death already. Which means it must have been! :)
>
> -PCP

Hi Pete,

Yes, I think you are right.  An MFM to IDE converter would be very
complex and complicated.  It is unlikely to exist in any form
especially as a hobbyist project.  The MFM ST506/ST412 very complex
and most controllers are unique to their drives.  Rarely can an MFM
harddrive formatted on one controller be successfully operated and
data read on a different manufacturer's controller.

I do not see how a ST506/ST412 to IDE interface is practical due to
the speed of processing the realtime data.  A Catweasel like device
for floppy drives generates a good amount of data on just plain floppy
disks at 125Kbps, 250Kbps, 500 Kbps signals.  Scaling up to ST506/
ST412 data rates would be quite dramatic -- 5MB or greater.  As Herb
said, this topic has been hashed out repeatedly here and elsewhere.

On the other hand, I believe a hobbyist SCSI-1 to IDE interface is
much more simple and practical since both interfaces are purely
digital and generally SCSI and IDE drives are intelligent devices.
The drives do a lot of the hard work such as the real time data
extraction and encoding themselves.  There is a homebrew hobbyists
SCSI-1 to IDE (and SD) project underway at vintage-computer.com
forums.  I designed a PCB for it a while back.

So far the SCSI to IDE/SD project has not really taken off although
there are about 5 or 6 builders with boards underway.  Hopefully we'll
see one come to pass in the near future.  The project is hobbyist in
nature so it uses common parts like Z80, RAM, ROM, and the Zilog
Z53C80 SCSI interface IC.  My hope is this project will be successful
before the last of the SCSI-1 drives disappear.  They were never as
common as IDE drives and their availability is steadily dropping off.

Thanks and have a nice day!

Andrew Lynch

On 02/25/2011 11:30 PM, lynchaj wrote:

> Yes, I think you are right.  An MFM to IDE converter would be very
> complex and complicated.  It is unlikely to exist in any form
> especially as a hobbyist project.  The MFM ST506/ST412 very complex
> and most controllers are unique to their drives.  Rarely can an MFM
> harddrive formatted on one controller be successfully operated and
> data read on a different manufacturer's controller.
>
> I do not see how a ST506/ST412 to IDE interface is practical due to
> the speed of processing the realtime data.  A Catweasel like device
> for floppy drives generates a good amount of data on just plain floppy
> disks at 125Kbps, 250Kbps, 500 Kbps signals.  Scaling up to ST506/
> ST412 data rates would be quite dramatic -- 5MB or greater.  As Herb
> said, this topic has been hashed out repeatedly here and elsewhere.

Andrew,

The nuts and bolts have already been done.  See the cctech mailing list for 
threads discussing Philip Pemberton's DiscFerret.  The PIC CPU he uses is fast 
enough to manage acquisition from an ST-506 MFM drive.  Discferret is designed 
for track-at-a-time reads/writes with transfers to a host PC over USB.  The 
core logic could certainly form the basis of an MFM-IDE adapter.  Not a 
beginner's project, but certainly feasible.