PDP-11 Controller needed to use a couple of RX33 drive

[Pete Turnbull]

On 19/06/2009 17:38, Johnny Billquist wrote:

> This was definitely in a 4-butten front panel BA23. However, I'm not 
> sure the front panel is the only place to blame.
> 
>> *Maybe* the fact that the Ready/Write Protect lines
>> for the 2nd drive are left floating, is somehow confusing
>> the RQDXn controller, but that seems doubtful.
> 
> The RQDX wasn't confused. It was the drives... When the RQDX started 
> writing to #0, the write circuitry of #1 also went active. Or if it was 
> writing to #1 which also activated the circuitry of #0.
> 
>> Those lines merely report status of the 2nd drive.
>> The Drive Select and Head Select lines don't go
>> anywhere near the front panel, and there are
>> separate signal lines for drives 0, 1, 2, and 3.
> 
> Exactly. Which is why I wonder if the front panel really is to blame for 
> the problem.

It's not to blame at all.  It has nothing to do with the drive selects.
  It might stop a drive working reliably, but it definitely cannot cause
two to be selected at the same time.  All it can do is tell the RQDX not
to use the drive.

> Hmm. As far as I can remember, there are no DS jumpers on the drives. If 
> you have an RQDXE, you have a bunch of jumpers on that in order to set 
> up which drive is which, and all that.

Not entirely.  It's intended to provide sufficient connectors and to 
route the some DS lines (etc) to the external unit.  That doesn't remove 
the need to set the DS jumpers on the drive(s) correctly.

> But the backplane of the BA23 don't have any jumpers, as far as I can 
> remember.

That's right, the backplane doesn't.

> And in this case, both drives were RD53s. Properly formatted, DEC 
> branded and everything.

Well, those drives definitely have Drive Select jumpers.  Look at the
legend on the PCB by contact numbers 32 to 26 on the 34-way connector.
They're the drive select lines, and you'll see they're labelled, and
they go to a set of jumpers right beside the connector position, next to
the terminator resistor pack.  If they were "as supplied by DEC" and the
person who installed them assumed the arrangement for a BA23 should be
the same as for a BA123 and didn't change one, they would both be set to
DS3, and so of course they both were selected at the same time.  User
error, pure and simple.

The microPDP-11 Maintenance manual says where they are, and moreover it 
says:

	Always place the first fixed-disk drive in port 0 (left mass
	storage slot of the enclosure containing the RQDX controller.

	Set the device select to DS3 on any fixed disk installed in port
	0 of the enclosure or expansion unit.

	Set the device select to DS4 on any fixed disk installed in port
	1 (right mass storage slot) of the enclosure or expansion unit.
	This rule also applies to any subsystem installed in a BA23
	enclosure.


-- 
Pete						Peter Turnbull
						Network Manager
						University of York