26 Jun
2012
26 Jun
'12
3:19 p.m.
----- Original Message:
Date: Tue, 26 Jun 2012 09:22:01 -0400
From: David Riley <fraveydank at gmail.com>
On Jun 26, 2012, at 5:43 AM, David Brownlee wrote:
> Does anyone know of any adaptors to fit a
"modern" drive (be it IDE,
> SCSI, ATA, CompactFlash etc) into a machine with an ST-506/412 interface?
That... would be quite the task. The ST-506 and
cousins directly output
the raw flux transitions as their data, so you'd need an emulator which
spit out the data as repeated cylinders, I believe. Trying to interpret
incoming data as low-level formatting would be another matter entirely.
You'd probably be better off emulating the interface to the drives (e.g.
emulate an MFM controller).
-------- Reply:
The main problem I see is that while the ST506/412 interface is standard,
there are quite a few different interfaces between an MFM controller and
the various systems (S100, DEC, PC etc.) and there are probably as many
different formats as there are controllers, so you'd have to emulate quite
a few different controllers.
But I've always wondered (and this may be a silly question): if a controller
in a 4.7MHz PC can handle the data flow, why would it be so difficult to
basically just connect another (pseudo-)HDC to the target HDC instead
of a drive, i.e. effectively connecting two HDCs together back to back?
Host MFM HDC
|| ||
Data/Control bus
|| ||
'Reverse' MFM pseudo-HDC & uC
|||||||
ATA bus or equiv.
|||||||
IDE/CF/etc. drive.
Assuming that the chips are available it doesn't look too difficult to
effectively clone a WD HDC and let it pretend to be a drive.
What's the obvious flaw that I'm missing?