Hi Geoff!
On Mon, 11 Aug 2014 17:14:54 -0700 (PDT), Geoff Wearmouth wrote:
> People often ask me. "Geoff: That Gosh Wonderful ROM - is it really so
> wonderful?". Yes it is.
Second that. It is really heartwarming to see, that an old hardware
platform such as the Sinclair ZX Spectrum still has a very active fan
community.
Some time ago, I fired up my Issue 2 Speccy (bank switched, external case,
keyboard, power supply, regulator, composite-out mod and whatnot), still
ran like a charm. OK, the inevitable color issues - no biggie. Even the
IF1 (first ROM revision) and microdrives didn't have defects. After
replacing some cartridge felted foam pads (dual side adhesive foam tape
did the trick), even external storage was online again.
However: Due to the dreaded print-through issues, none of the cartridges
was readable without errors. So I wrote a tool for RAW sector extraction
and PC interfac (beta stage, available for download at
ftp://ftp.worldofspectrum.org/pub/sinclair/tools/pc/mdv2img.zip) that
confirmed the problem.
Perhaps some data recovery expert can help and recombine the raw sector
image (redundant passes possible) with existing HDCHK (header checksum
error), DESCHK (record descriptor checksum error - that's what links
sectors together) and DCHK (data checksum error - e. g. corrupt sector
payload data) to something that actually works.
However, I doubt that this will succeed with the current Microdrive
hardware. Why? The ULA2G007 has an x200/x8 amplifier and Schmitt trigger
already built-in (see
http://www.1000bit.it/support/manuali/sinclair/ql/SINCLAIR%20QL%20SERVICE%20MANUAL.pdf,
page 63, diagram D9 - this is the QL service manual, but that doesn't
matter much, since the QL has the same Microdrives that you could connect
to the Speccy - via the IF-1, that is).
The problem is the Schmitt trigger. It will work well with clean signals,
but in our case the signal consists of two (or more) overlayed rather
sinusodial signals having about 1:2-1:5 amplitude ratio and varying phase
(not uncommon with print-through phenomena). In that environment, the A/D
conversion process fails miserably. And there's no chance getting around
that (e. g. by tweaking an amp, adding noise suppression stuff) since
everything is "hardcoded" in the ULA.
That said: Thanks for sharing your efforts with the community. Much
appreciated. Keep up the good work!
Volker
--
@: I N F O at B A R T H E L D dot N E T
3W:
www.bartheld.net