Alt-BEAM Archive
Message #11609
To: "'beam@sgiblab.sgi.com'" beam@sgiblab.sgi.com
From: Wilf Rigter Wilf.Rigter@powertech.bc.ca
Date: Sat, 11 Mar 2000 12:43:30 -0800
Subject: [alt-beam] Re: Diodes in microcores
Hi Jeff,
Welcome and a pleasure to read about your Nv circuit observations.
If you use LEDs to indicate the active low output state of each process, you
probably noticed that the 4Nv microcore starts up saturated with 2 active
processes shown by the following LED sequence:
1 0 1 0 (On Off On Off)
0 1 0 1
1 0 1 0
etc
and a micro core with PNC starts as:
1 0 1 0
1 0 0 1
1 0 0 0
stop
until the PNC times out and continues:
0 1 0 0
0 0 1 0
0 0 0 1
1 0 0 0
0 1 0 0
etc
If you have motors connected to the microcore perhaps using a 245 buffer or
h-bridges, then the motor connected to the PNC controlled Nv will be turning
in one direction until the PNC times out. The usual solution is to add a
connection from the PNC to the 245 chip enable pin to tristate the buffer
and disable the motors until the PNC times out. H-bridges are more difficult
and some will self destruct if both inputs are active! Microcores are now
probably less popular because of the unpleasant experiences of smoking
h-bridges. The 2 diode PNC free solution eliminates that problem altogether
and who knows how Beam would have evolved if that solution had be been found
early on.
I like your analogy of energy states of neural networks and I must think
more deeply about that one. But I wonder if the saturation isn't the lower
state and a single process, the higher state. I say this because the number
of different output patterns generated with one process is 4 but with 2
processes (saturation) the number of different output patterns is only 2.
In fact, if the motors are connected in the normal way from the output of Nv
1 to 3 and Nv 2 to 4, when saturation occurs, the motors do nothing and no
motor current flows because the levels on each the motor terminal are always
the same polarity.
So reading your observations made me curious since the behaviour you
described is a bit different from what I expected but hey, I am happiest
when I make a counter-intuitive discovery and you never know what new ideas
may pop up next.
regards
wilf
BTW I'm too fumble-fingered for freeforming any but the simplest circuits
myself but others have perfected it to a fine art!
Home