S/H for logic signals?

[torsig1967]

I'm trying to reduce processor needed for logic signals in NM G1.

Is there som way to realize these objects with logic gates and other CPU-effective stuff? They only have to work with logic signals, not audio signals.

- Sample/Hold
- Controllable 2 to 1 switch
- Controllable 1 to 2 switch
(With controllable I mean they should change state when controlsignal changes from 0 to 1 and vice versa)

For the moment I'm using the built in audio S/H, X-fade and pan and they-re not at all CPU-effective. Especially not S/H which needs 1%.

[Kassen]

S&H can be done with a OR gate. Feed the output back on a input. As as you now put a 1 on the other input it'll stay high forever. Put a AND gate inside of the feedback loop and feed the other input of that with a logic inverter. Now putting a 1 on that inverter will break the chain and get the OR to output 0 again.

1 to 2; feed your signal into two seperate AND gates. On the other input of the first you put your controll signal, on the other input of the second a inverted version of it.

For the 2 to 1 you can use the same trick of two AND gates except that there their inputs will be the different signals and you can mix them with a OR.

These tricks are typically a bit cheaper on the cpu but a bit harder on the zero page. Also remember that complex logic chains may suffer from calculation order problems and that could be fixed by using the faster processing of the x-fade and so on.

[torsig1967]

Thanks, switches work nicely but I cannot see that the S/H works correctly. Or maybe I didn't understand how you mean. I connected it like in the enclosed picture and it will not work as a proper S/H. Logic 1 on the input will produce logic 1 on the output even before the trig signal. And there's a zeroing issue when output is 1 when the trig arrives.

---

not_sh.gif
Description:
Filesize:	9.98 KB
Viewed:	3527 Time(s)

---