11/45 progress

[Tony Duell]

> Reattaching the front panel was interesting, a previous person had 
> overtorqued one of the nylon screws too tight and it had broken off inside 
> the standoff. I one-upped them by doing the same thing. I tried drilling out 
> the two nylon screws to no avail. I decided to try to melt them out. Who 
> would have thunk of using a blowtorch on an 11/45 front panel ;) Interesting 
> physics... within about 10 seconds of being in the flame... I'm guessing the 
> trapped air inside the standoff expanded, because the screw unthreaded 
> itself and backed all the way out. Cool.

I hope you unscrewed the stand-off from, the CPU chassis first...

> 
> The next steps will be power supply checkout and front panel. It has been 
> suggested that I power it up with a minimal card set. That's fine and what I 
> would normally do; however, I was wondering if it was ok to pull every card 
> including the cpu set and power it up, just to check the voltages on the 
> backplane.

You can run this PSU without load. The voltages will be high, but there 
will be no damage. What I did with my 11/45 was run the PSU without load, 
bring all the voltages to the right values with the presets on the PSU 
bricks, then put the cards in and re-adjust them back to the right values.

There are several +5V lines on this machine, from separate PSU bricks 
(common ground, of course). I think the CPU uses one (or maybe 2), the 
floating point (and maybe MMU) another, Unibus expansion another one (or 
again maybe 2). You need to check all of them (see the CPU prints, the 
PSU distribution wiring diagram, which is different between early and 
late machines, for detais as to what goes where).

> 
> Speaking of which, how can one test the voltages without an extender card? 
> Maybe hunt down the right pins on the backside of the backplane?

Asuming nobody has messed around with the power harness (this is not the 
same as _modifying_ the harness, which was quite common to get more 
current on the +5V line to the Unibus, etc, if you wern't using the 
Fastbus memory), then what you're worried about is a defective power 
brick putting out too high a voltage, you're not worried about the +5 and 
+15V lines being crossed. In which case you can check the PSU outputs on 
the connectors to the bricks and on one of the connectors for the PSU 
control board (one of the 3 connectors on the rear of the transformer 
housing part of the PSU).

> 
> Then I've got a question about termination, but I'm still going through some 
> documentation to see if I can figure that out myself.

Simple. An M930 (or I guess one of the boot/terminator cards) in the A/B 
slots of the front slot (1) of the CPU backplane. A similar terminator 
card in the A/B slot of the Unibus Out of the last Unibus backplane (if 
just the CPU, put it in slot 28).

A couple more things. Unless you are soing something seriously odd, put 
an M9200 jumper (like an M920, but only 1/2" spacing) in A/B of slots 26 
and 27 of the CPU backplane to strap the 2 buses together.

The grant chain (or at least one of the Bus Grant lines) goes via slot C 
of slot 1 of the CPU backplane, which is wired for a KW11-L line time 
clock card. If you don't have this card installed, the grant chain is 
completed by a wire-wrapped jumper on the backplane (which you cut/remove 
if you fit the KW11-L). You probably have this card, but if you remove it 
for testing and then have interrupt problems, you'll know why.

-tony