DG S/130 progress
cube1 at charter.net
Tue Oct 13 19:04:22 CDT 2015
On 10/13/2015 5:08 PM, Jay West wrote:
> Pictures of the S/130 system I'm building up are at
> I think the last I posted here... the cpu was up and running, that was the
> first piece I refurbished. I took everything out of the rack and cleaned
> that up well, and the cpu has been remounted there.
> Yesterday I mounted the 6030 dual 8" floppy drive. It was squeaky clean
> inside and at power up everything does what it should. However, I can't
> proceed further with it as I have no media. The docs say it uses hard sector
> 8" floppies, which I believe I can find. However, it specifies 8
> sectors/track hard sector floppies - and those I've never heard of. Maybe DG
> made them special just for their gear, or, maybe the drive or controller or
> software was smart enough to expect the more common 32 hard sector floppies
> and knew to only pay attention to every 4th sector hole ;)
I see that sector count (0-7) in the 6030 technical reference.
But all of the HS floppies I have for DG are 32 sectors - and many of
those are software distribution floppies for Novas, appropriate to the
6030. They are all labelled DG Diskette (as opposed to /2, /3 or /4),
as per the illustration in the 6030 manual.
The manual specifies a max sector access time of 166ms, and a min of
1.2ms. 166ms implies a rotational speed of 360 rpm (6 revolutions per
second, or 166ms).
Divide that by 8, and we get sectors about 20ms apart (8 sectors), with
an extra index pulse 1/2 of that, or 10ms.
For 32 sectors they would be at 5.1ms and the extra index pulse would be
1/2 of the 32 sector pulse, or 2.6ms.
Looking at the schematic I have, 001-000797-10 (rev 10), sheet 2, bottom
left quadrant, the index pulse is fed into a one-shot with a time
constant noted of 3.9ms to generate the index pulse - which would be
pretty darn close to 1/2 of the way in between 2.6ms and 5.1ms.
The sector pulse is derived from a counter, U63. The counter has an
input signal labelled "DIV 4" to pins 5 and 6.
Also (this would have saved time had I seen it right away) there is a
timing diagram, 001-000893-00 that shows 5.2ms between sector pulses,
and 2.6ms for the extra index pulse, with pulses being 1.1ms in length
and also shows the 3.9ms one-shot timing.
That cinches it: the design is for 32 sector floppies.
More to follow related to my hardware and software inventory, but I just
got called to dinner.
More information about the cctech