At 01:23 PM 12/28/00 -0500, you wrote:
>Following Isildur's earlier mention of a ``custom'' MicroVAX.....
>
>I may dredge the MV3300 out of storage this weekend and see if I can
>run up some spare MVII boards in it.  For the sake of discussion,
>and so I don't inadvertently blow something into the great silicon
>bit-bucket in the sky, can someone refresh my memory about using
>MVII boards in the BA213 cabinet?

I did this to test some MVII boards and to look at slow boot times. You 
will need the cabinet kit for the KA650 CPU. It has two connectors (one for 
console and one for control panel) as opposed to the KA640 cab kit that has 
only one connecting cable. There are also s-box handles for DELQA boards 
but since I was testing a DEQNA I just hung the cable out the front. The 
system used the SCSI controller and the TQK70 that were in the cabinet. 
(I'd like to test the TQK50 but I can't easily wire up a TK50 to it.

>I have a TKQ70 and TK70 drive, originally from the MV3300.

You'll probably want to put these back in? (When you say "from" did you 
actually remove them? If not they will work fine with the MV2)

>I have a SQ706 scsi mscp board and can wrangle up some
>1 gig drives off another VAX or DECstationthingie.
>I could also set a storage expansion box on top of it
>or some kind of external scsi box if needed, and cabling
>was a problem.  Basically, it is just to lash it up to
>running capability.

Presumably the SQ706 will just work but I don't have one yet so I haven't 
tried it.

>I have a DEQNA ethernet card (since the ka640 bits are not
>supported.  Will it pose problems on the MV3300 bus?

No just make sure everything is plugged into the "top" half of the bus, 
there are no Qbus slots in the lower half. Give the BA215 you may run out 
of slots.

>Somewhere I thought I ran across some DEQNA warnings on
>later Qbus machines, but I don't remember where, offhand.

The faster processors can overwhelm a poor DEQNA.

>I have the original KA630 cpu and ram set from another MVII,
>but I understand those WONT work in the BA213 cabinet, right?

I've got empirical evidence that they do work.

>Now, the questions arise as to how to properly set up the
>boards into the BA213.  Granted, cabling is a problem, since
>I don't have the proper front panel bits usual to the MVIII
>kind of line.  But, I have the MVII cabling and patchpanel
>bits, and could affix them to the BA215 howsumevertomakeitwork.
>The scsi does not require any patchpanels, and the drives don't
>use the early mfm/esdi pushbutton panels.
>
>Are there any particular gotchas I should be aware of in trying
>this hybridization/bastardization?  The end goal is to get another
>MVII/MVIII sort of thingie up sufficient to test some NetBSD bits
>while my other MVII is 4.3BSD'ing.

The only one being where you can place the boards. This will limit you to 6 
boards total and with MVII + Memory taking up 2, DEQNA + tape + disk with 
3, that leaves you 1 to play with. So if you have two memory boards it will 
be full.

>Are there any particular board orderings I should use?

 From right to left:
         CPU
         Mem
         Mem (opt)
         DEQNA
         DISK
         TAPE

>My expectation is that the TKQ70 goes first, followed
>by the DEQNA, followe by the SQ706A.  Is that satisfactory?

The TQK70 should go last, its very good about waiting for the bus, the 
DEQNA on the other hand has no buffers to speak of so it needs the but in a 
hurry when it gets a packet.

>Is it correct to assume that the board lineup is straight across
>the top A/B ports of the bus, and NOT serpentine?

Yes.

>That is all I can think of, offhand.
>Thanks for any suggestions/insights.....

It will be fun. I may have a spare KA650 cab kit if you decide you want to 
do this on a more permanent basis :-)

--Chuck