8 bit SBC with floppy controller for disk archiving [was Xerox820system disks in Teledisk format]

[Allison]

>
>Subject: RE: 8 bit SBC with floppy controller for disk archiving [was Xerox820system disks in Teledisk format]
>   From: "Chuck Guzis" <cclist at sydex.com>
>   Date: Mon, 11 May 2009 20:12:32 -0700
>     To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>On 11 May 2009 at 22:26, Andrew Lynch wrote:
>
>> Hi Chuck!  Thanks!  By extensions, I mean the Disk IO has a regular
>> Shugart mini-disk interface with all the usual signals and options. 
>> It also has the "feature" interface which imports/exports the rest of
>> the NEC765/i8272 signals necessary for 8" and other drive support.  An
>> 8" drive cable with what signals are needed for that drive could be
>> fabricated using the Shugart mini-disk interface plus whatever
>> signals/options are needed from the feature connector.  The 8" drive
>> cable would be custom to Disk IO and that drive but there are multiple
>> 8" drive interfaces so that seemed to me to be the most flexible way
>> to provide support in a limited PCB space format.
>
>Hi Andrew,
>
>Good work!
>
>I did check out your schematic (I think)--FDC_A.jpg and FDC_B.jpg and 
>I don't see the "special feature" connector anywhere.
>
>Here are my comments on what I see, they're only speculation and 
>opinion, so please don't take them for anything more than that.  I'm 
>certain that you had good reasons for doing what you did the way you 
>did it!
>
>One of the problems with using a 4MHz Z80 on a raw data stream is 
>that the Z80A isn't fast enough to handle it.  While a uC might do 
>that, the simplest seems to be the Catweasel's use of a 128K time-
>domain memory.  Leave it to the host processor to crunch the data 
>later.  One could well more easily hook up a uC to the raw disk 
>interface, don't you think?

To do NS*, Victor and Heath hard sector and other odd formats would
be outside the range of Z80 even at 10mhz.  Most of those had unique
external hardware to assist.


>How does one derive the RWC/TG43 signal that some 8" drives require?  
>If there's an 8" connector, why is HDL not brought to it?

The 765 does support that its' implmentation that determines if its
brought to a connector.


>One thing that I'm not certain of is the use of the US0 and US1 pins 
>for select.  The PC uses a separate register for select and leaves 
>US0 and US1 NC.  I suspect that this bears on some aspect of 5.25" 
>operation and the 765's  polling mode, but I'm not certain. 

765 has unit select and head select.  Typically when it's polling unit
select is active for the four drive poll (ready line).  When a drive is
selected for read or wrie head (HDL) is enabled with internal timers
and a read or write will take place.  The PC uses a seperate register
for that and side steps that feature.


>Would it be better to multiplex the READY signal to a set of jumpers 
>that allows one to select either the drive's own ready (all 8" drives 
>and most 5.25" have this) or a tied-high READY?

The full out designs for 765 do that but most simplify to cust chip/cost
and usually that is the first to go.

>FWIW, the Victor 9000 disk format is variable datarate (or CLV, 
>depending on how you want to look at it). 

It's also very unique to Victor and even when a new thing like so many other
non portable formats caused the user great pain in importing new software
available to those that had more generic formats. 


Allison


>Best regards as always,
>Chuck
>
> 
>