19 Jun
2011
19 Jun
'11
11:03 a.m.
I have early calculators where flip-flops are built
from a single
physical OR gate.
Now that's one I've not seen...
The output feeds back to one input so the gate will
latch-up on itself.
The other inputs to the OR gate are SET inputs, sending one true will
latch true.
The gate is OC, so RESET inputs are created with simply diodes or
direct connections wire-ANDed to the output/loopback-input. Sending
such an input false will force the output false and it will latch in
that state.
So it really has an AND gate too, albeint a wired AND of open-collector
signals. Makes sense.
I have an instrument which has flip-foops made from cross-coupled
open-collector inverters. The operation si obvious, it's stable in either
state nd you filp states by pulling one or other of the outputs,
What makes it slightly odd is that these flip-flops are linked to
backplane signals wwhich are effectively bidirections. For example, there
is a signle signals for 'Print Command/Complete'. Some boards inthe
backplane will briefly pull that signal low, it will be latched by such a
flip-flop on the pritner itnerface board, and thus the backplane signal
will remain low until the pritner interface board clears the flip-flop
(which it dos when the pritner has accepted the character). I've not seen
that trick done anywhere else.
-tony