Somewhat random QBus backplane / chassis question

[Josh Dersch]

Hi all --

I've got an 11/03 backplane here (designated as an "11/03 EA" originally 
but marked as "Model changed from EA to AA") inside a small chassis 
w/power supply, model "OH780-A."  (So it was at some point in its life 
the front-end to a VAX-11/780.)  As far as I can tell, it has a standard 
H9270 backplane (4-slot, quad width) in it.

The chassis appears to work fine when I have an actual LSI-11 in there 
(I've been experimenting with an 11/23, but I do have an 11/03 boardset 
that works in it).

What I want to do with the chassis is use it with a UNIBUS->QBus bridge 
I have, as a Qbus backplane for a SCSI controller (and maybe other 
things) for my VAX-11/750.  I have encountered an oddity in doing so, 
and I've looked at the documentation and I remain puzzled.  The oddity 
is this:  If I don't have an LSI-11 CPU board in place in the chassis, 
the power supply won't come up properly.  (The +5V appears to pulse, the 
kind of behavior I'd expect with a shorted or heavily loaded supply rail 
somewhere).

Here's a basic rundown of what I've observed:

- If I put an 11/23 or 11/03 CPU in, everything works fine.

- Doesn't matter what slot the CPU's in either, as long as there's a CPU 
board somewhere in the thing, all's well.

- Doesn't matter how heavily or lightly loaded the backplane is; I've 
tried it with just a QBus SCSI controller, I've tried it with the 
backplane stuffed full of boards.

Since I want to use this as a simple QBus backplane, I don't want to 
have a CPU board in place.  I haven't seen this behavior before (but 
then, I can't recall trying to run a Qbus backplane without a CPU in it 
either).  I thought I knew QBus fairly well, so I assume I'm overlooking 
something obvious.  Or perhaps there is a fault here with the chassis, 
somehow.  I figure someone here will have the answer immediately and 
I'll slap my forehead...

Thanks as always,

Josh