M8650 async board problem - test-drb@ccmp.vtda.org

22 Dec 2001

I've fixed it.

On Dec 20, 19:36, Tony Duell wrote:
...
  Pete wrote: 
...
  > Now I'm really puzzled.  After spending the
evening playing with my
 > serial board, and armed with the pinouts this time, I still have an
 > interrupt problem. 
...
  Be warned DEC can be confusing here. For example on
the schematics, pin 8
 of the FF is shown as Q and pin 9 as Q/. But that's because the D input
 on pin 12 has an inverting 'bubble'. And of course the Set and Reset pins
 are therefore the other way round to what you'd expect. 
The confusion comes mostly from the Maintenance Manual.  Thanks to David
Gesswien for scanning the KL8-E printset, and Tony's description of the
pinout labelling, I worked out what should be happening.

...
  That's the problem, I think. DEC's data bus is
active low. Or that's how
 it appears to me. 
Indeed, and that's what I expected.  The Handbook doesn't show the data bus
signals as active low, whereas it does secifically show 'L' after other
active-low signals.  Nevertheless, like most other open-collector type
busses, the Omnibus is active low.

...
  > Should I just swap the connections to the SET and
CLR, and Q and not-Q on
...
   the 7474?  Or
am I missing something? 
 No (!). It worked once as it is now, so it can work again. Find out if
 you can ever make the Q/ output of that FF low, and if not, why not. Are
 you getting a pulse on the clock pin when you execute a KIE instruction?
 What is the D pin doing? 
They all seemed to be doing something, just not necessarily the *right*
something, when I first mailed the list.  I wasn't sure it actually *had*
worked properly; I wondered if there could be an ECO missing.  Certainly it
hadn't worked for a long time, and presumably that was why someone had cut
the INT RQST L track a long time ago.

However the fault, for those who're still reading, wasn't really the M8650
itself.  There were two problems.  At some point, I had unplugged the
serial cable from the Berg connector, so there was noise on the serial
input, which upset the diagnostics.  The second problem was that the PSU
+5V was down to about 4.5V, which isn't great for TTL.  I'd checked the
voltage when I reassembled the machine, but it didn't have a full
complement of boards in it at that time, and the additional load made it
droop due to several not-too-tight connections between PSU, power cables,
and backplane.  I've since cleaned them up extra carefully, and tightened
the connections, and readjusted the PSU to give 4.9V on the front backplane
and just over 5V on the rear one (the difference is due to different loads
on the two parts).

Now KIE with AC11=0 turns interrupts off, as does CAF and the CLEAR switch,
and KIE with AC11=1 turns them on, and it passes all the diagnostics.

Oh, and FOCAL 69 runs fine :-)  Now onward to OS/8 and some RX01s...

Thanks to everyone whose suggestions or websites helped me get this far:
David, Kevin, Tony, Allison, Ethan, Doug, Zane, the other Kevin, and Nabil.
 And probably others I've forgotten.

-- 
Pete						Peter Turnbull
						Network Manager
						University of York