Sam Goldwasser wrote:
"cooldude" writes:

Hi,

1) Is there any equivalent to monostable in fully digital electronics
(logic gates)? I am trying to trigger a pulse (a set width) with a very
narrow recovery time (in the range of 100ns or less) before the next.
74LS221 seems to work although getting weird pulses (very short
retrigger pulses) occasionally.

This is not a problem with the device but rather power supply decoupling,
glitches on the clock signal, a race condition, or noise on the wiring.
Monostables work reliably when properly fed.


Sam's right. Any noise on the power buses is going to be bad news for
the monostable. That's because a monostable works by comparing a
rising ramp voltage with some logic threshold. There's no problem at
first, because the ramp is down near zero volts and the threshold is
like 1/2 to 2/3 Vcc, But as the ramp rises, the monostable keeps
trying to compare 1, 2, 2,.5, 2.6, 2.7 volts with the current
threshold. As the time interval draws to a close, it's trying to
compare smaller and smaller differences. Now the other fully-digital
chips won't be bothered by a volt of noise, but this comparator will
jitter all over the place, in proportion to the jitter on Vcc.

Try putting a 100 ohm resistor in series with the +5 volts going to the
monostable and bypass that to ground with a 10uF capacitor PLUS a 0.1uF
ceramic. Make sure the timing resistor and capacitor are fed off this
decoupled Vcc source too!..