ImageDisk and some 8" images posted

Dave Dunfield dave04a at
Wed Aug 3 09:32:35 CDT 2005

Hi Allison,

>>Unfortunately it's looking like mixed sector sizes within one track are
>>not within the capabilities of the 765 to write - although I could probably
>>READ them OK (and with a documented image file format, you would at least
>>have access to the data so you could try to find some other way to recreate
>>the disks if you had to).
>Dave there is a trick.
>Looks at FCOPY at Gaby's (unoffical CPM archive).  Tim and I developed it
>for a project and it's not a generic imager.  However we found a few things
>out and one is that most 765 implmentations will lock up if in the worng 
>mode (SD or DD) initially. So to do mixed sector you can't write.  However 
>reading is possible if you first try the mode that doesnt lock first then

Thanks for the info - I'll check it out when I make the changes to support
mixed density reading ... (which may be a while).

>In the end you will be recreating the catsweasel. If you really want to
>go to the next level starty with a PC fast enough and skip the whole FDC
>chip thing.  The next level is a data speperator logic and a shift
>register that can be read at the bit rate by the PC and do the needed
>pattern regognition (look for sync, address marks and the like), and then
>build a literally imaged disk and that can be anaylysed.

I've thought about this - I can do 500khz samples on a PC parallel port,
and with some simple hardware I could get 8x this - but it seems dodgey,
and unless I disable all interrupts (which has it's own set of problems
including losing HD access) there would be excessive real-time jitter.

I've also thought about dropping in something line a 90Mhz (or faster)
ARM - which would give me a much higher sampling speed and better real-
time control than a PC ... but this gets tough to build...

The 2793 would be easy to build, and should handle most soft-sector disks.
The "read track" and "write track" functions of the WD chip allow near raw-
access to the track data.

But these questions are part of the reason I've not done anything toward
a hardware solution so far.

dave04a (at)    Dave Dunfield
dunfield (dot)  Firmware development services & tools:
com             Collector of vintage computing equipment:

More information about the cctech mailing list