On 07/11/2008 21:07, Tony Duell wrote:
I assume you realise that an RX50 is just a pair of 80
cylinder single
head floppy drives with a common disk spindle and common head stepper.
The 34 pin interface cable is essentially the standard pinout. The only
trap for the unwary (if I am reading the schematic correctly) is that if
side select is asseted (this would select the other side of a double-head
drive, of course), then the RX50 _disables_ its track 00 output. I think
at least one machine used that feature to detect a real RX50.
I found that out the many years ago, trying to get non-DEC drives to
work on several variants of the RQDX1 and RQDX2. I came up with a
little 1-LSTTL circuit to kludge the effect. I've a feeling that may be
how an RQDX3 tells the difference between RX50 and RX33 drives. ISTR
there's also something about the index pulse that meant some drives
didn't work, something up the way the controller checks the drive is up
to speed. Ah, yes, see below.
Other differences:
1) The RX50 doesn't use or need the HeadLoad/InUse signal on pin 4.
2) The RX50 will only respond to step pulses if it is selected *and*
Motor On is valid; many SA400-type drives will step even if MotorOn is
de-asserted.
3) The Ready signal on Pin 34 is asserted whenever the drive is
selected, has a disk in it, and the door is closed. Most drives only
assert Ready when the drive is up to speed (often by timing the interval
between index pulses).
The inherent delay on non-genuine-RX50s means that some RQDX controllers
don't see the drive is Ready immediately, and take it offline.
--
Pete Peter Turnbull
Network Manager
University of York