<Interrupt latency is determined by the cards slot position. Lower number,
<lower latency. Some can tolerate that (DZ11's for example) some can't (som
<disk controllers)
<
<>I might expect the QD696 to be in slot 4 top, with bus arbitrator in botto
<
<That works, although on my BA23 I've got the DEQNA first (actually I have 
<DELQA but its functionally the same card) then the Dilog 696 in next, then
<the tape controller third. (slots 4, 5, and 6 in my picture), then a bus
<grant, and then the DHV-11 (quad wide).

Most disk card can be down the grant chain as thy have silos/buffers and 
interrupt latency is usually not a problem.

the priority is, Higest to lowest...

 LPV11
 DLVj11 (or any DL type serial)
 Deqna/Delqa

 VCB02 base video can be after the above or after below.
 VCB02 (additional 4plane modules)

 DHV11
 DZV-11(M/N)
 DMV11
 RQDX3
 KDA50

Note RQDX1 and some RQDX2s MUST BE LAST as they dont have bus grant 
continuation.

<>Maybe VIK/QTO scsi (is it tape or tape/HD capable?) in slot 6 with
<
<This is TAPE only (Q = Q-bus, TO = Tape Only)

This can be way down the list as it's buffered.

<>earlier, but someone removed the video to make it a server of some
<>sort, perhaps, in MVII rather than VSII configuration?
<
<Sure sounds like someone converted it from a VSII/RC to a MV II. I was at 
<DEC resller once who had a variety of plastic plates for the front of thes
<things :-)

Sounds like that thing went through a whole load of chages over the years.

<>What is the prom command for low-level formatting the drive, if needed?
<>Test 75?

No.  Only 3100s have that.  the Roms for MicrovaxII are limited to:

*B(oot)
 X(binload/unload)
*C(ontinue) usually from a break/halt
*D(eposit) 
*E(xamine)
 F(ind)  
*I(nitialize)
 H(alt)
 S(tart)
 R(epeat) command
 T(est) (valid test numbers are 3-7,B... same as power on tests)
*U(njam) (issue bus reset)

* most often used.

Allison