24 May
2000
24 May
'00
3:22 p.m.
Are you using a controller that requires the external "data separator" on
the drive? If you use it with a controller that expects the raw data you'll
get something that works part of the time. One might interpret that as
being pattern-sensitive, since it is.
I'm not sure you need to fiddle with this data separator much, but if you
take into consideration that what most folks call a "data separator" is
really a clock extractor, which is little else than a missing clock detector
and fills in the missing clocks for the benefit of the data extraction logic
in the controller.
Why not contact me off-list and I'll look up what I can from the drive
documents.
What troubles me is that I don't know what kind of controller you're using,
hence I dont know that this mechanism on the drive is even required for it.
Are you sure you need this crude bit of circuitry?
Dick
----- Original Message -----
From: Dwight Elvey <elvey(a)hal.com>
To: <classiccmp(a)classiccmp.org>
Sent: Wednesday, May 24, 2000 12:19 PM
Subject: Shugart drive question
Hi
I'm still working on my NIC-80. I'm having a data pattern
sensitive problem. I'm not yet sure if it is in the SA900
drive or in the controller. When looking at the drive,
there is a clock/data separator circuit. It uses two one shots.
One is for the next clock cycle after a one was read and
one is for the next clock cycle if there was no one read.
I understand the issues of this shifting of the read
data but I don't know how much these two one shots need
to be adjusted. The SA800 TTL board also has the same circuit.
I was hoping that someone with an old drive could put
a 'scope on these two one shots and determine the nominal
pulse widths they see? The test points are labled the same
on both the SA900 and SA800 series drives as TP21 and TP24.
They are negative going pulses.
Dwight