Re: [sdiy] Binary state change detector

[Roman]

You don't need 42 to latch, a RC will do.
connect each input directly to A input of hc85
and via 1k to B input, which has 1n cap to ground.
Every changing bit will cause mismatch on A/B
inputs for more or less the time constant of RC.
This way you save 1 IC and have full control
on pulse width. If you make it as long as intended
one shot, you can even leave that part off.

Roman

---- Wiadomość Oryginalna ----
Od: Peter Grenader <peter at buzzclick-music.com>
Do: Synth DIY <synth-diy at dropmix.xs4all.nl>
Data: Wed, 08 Oct 2003 07:20:36 -0700
Temat: [sdiy] Binary state change detector

>I am looking for a circuit that detects when the state (number) of a  four
>bit binary word changes state.
>
>I've been putzing around with the idea of using a 74C85 in conjunction with
>a 4242 (using the 42 to latch the the last occurrence of change and feed it
>back into the C85 for the next comparison, and then running that output
>through a oneshot to keep it from disappearing as soon as it is sensed, but
>to this I have a question:
>
>How short of a incoming pulse does a oneshot require in order to react?  For
>instance, another perfect example is a sequencer reset pulse coming from one
>of it's stage gate outputs.  As soon as it goes high, the reset function
>occurs and the gate is no longer high.  Will that pulse width, as short as
>it is, be long enough for a oneshot to react?  Although this isn't my
>application, its the same idea and if it's long enough, me thinks I've got
>the circuit I need here.
>
>If not, I'm totally screwed and am asking the masters here if they have any
>other suggestions.
>
>thanks in advance,
>
>Peter
>
>