substitute

How to generate "words" for CMOS demux?
« on: June 02, 2020, 10:53:54 PM »
I have an idea for a riff on a sequential switch and the CD4051B basically does what I would want.
http://www.ti.com/lit/ds/symlink/cd4051b.pdf?HQS=TI-null-null-mousermode-df-pf-null-wwe&ts=1591120375019

So, on page 1 I'm looking at the top configuration (1 com  to 8 outs). What I'd like to do is drive it with one pulse train and rotate through  the 8 outputs then reset and repeat like a decade counter. However, in order to toggle through the various outputs I need to generate what I guess would be called a 3 bit word on A, B, C  for each step (i'm fuzzy on this stuff so apologies if I'm botching the nomenclature).  So, how can I generate these words from a single pulse without a ton of complicated logic? There's gotta be an IC that does that right?

Thanks


squarewave

Re: How to generate "words" for CMOS demux?
« Reply #1 on: June 03, 2020, 12:11:46 AM »
It's just a binary input. So 000 is 0, 010 is 2, 101 is 5, etc. No IC is required but it's a digital input so unless you're using a binary encoder or something else, the alternative is a microcontroller. With a mirco you could just make a point of using pins on the micro that are 3 consecutive bits of the same register and then just write a number to the register directly (aka direct port manipulation in Arduino speak) and voila, that gate is connected.

However, you do know that that part is pretty lo-fi right? Distortion is like 0.1% under the best conditions. If this is for some synth thing then maybe you don't care but if you used something like ADG1408 (same deal 1:8 with a parallel 3 bit binary input) driven from a low impedance and loaded with a high impedance, it would probably be good enough that it would be difficult to measure the distortion.

Re: How to generate "words" for CMOS demux?
« Reply #2 on: June 03, 2020, 06:20:49 AM »
Hello

Search for mod 8 counter, or 3 bit counter.
Should be possible with 3 flip flop. or BCD counter

I just have a look at 74LS9x and the 93 (full 4 bit counter) seem to do modulo 8 by Q3 as reset

Best
Zam

substitute

Re: How to generate "words" for CMOS demux?
« Reply #3 on: June 03, 2020, 07:50:13 AM »
Quote
However, you do know that that part is pretty lo-fi right? Distortion is like 0.1% under the best conditions. If this is for some synth thing then maybe you don't care but if you used something like ADG1408 (same deal 1:8 with a parallel 3 bit binary input) driven from a low impedance and loaded with a high impedance, it would probably be good enough that it would be difficult to measure the distortion.

I did not realize that and it is indeed for synth stuff. What attracted me to the part was that it could switch 20v p to p signals, but I'll look into the suggested parts, thanks.

Quote
Hello

Search for mod 8 counter, or 3 bit counter.
Should be possible with 3 flip flop. or BCD counter

I just have a look at 74LS9x and the 93 (full 4 bit counter) seem to do modulo 8 by Q3 as reset

Best
Zam

Thanks Zam.

I found this article from an old issue of Polyphony, which I could probably tweak to do what I want. http://www.muzines.co.uk/articles/build-a-quad-sequential-switch/5287


Re: How to generate "words" for CMOS demux?
« Reply #4 on: June 03, 2020, 09:05:43 AM »
CD4017 won't work, it's a pulse count (each output high only for one time clock)

look at 4510, 4518/20 , 4553...

there is various CMOS counter, DCB, binary etc ...
pick one with reset (to be able to count 8 and not more)

you should be able to use just one push button , the counter and the switch...
If that's the request you have ?

Best
Zam

squarewave

Re: How to generate "words" for CMOS demux?
« Reply #5 on: June 03, 2020, 09:23:29 AM »
Ah, I see. No problem. Yeah, schmitt trigger into BCD counter into 3 bit binary coded analog gate would do it. However it seems biasing is a little different. Modern analog gates are dual supply and signals are biased in the middle but they have a ground reference used by digital inputs. So the schmitt trigger should make 0-5 and not +-V. Or maybe you don't need the schmitt trigger at all but there should be some kind of protection for the analog gate so I would use it. It also gives you some control over triggering levels.

Note that AD1408 is a tiny chip (as are many of the analog gate chips) but very possible to solder onto a "surfboard" by removing excess solder with a braid and some flux. Then clean up the mess with 99% isopropyl alcohol and a q-tip.

substitute

Re: How to generate "words" for CMOS demux?
« Reply #6 on: June 03, 2020, 01:42:30 PM »
Sifting through these replies now and studying data sheets, will respond directly in a bit. I made a very simplified block diagram of what I'm looking to do.




L´Andratté

Re: How to generate "words" for CMOS demux?
« Reply #7 on: June 03, 2020, 06:09:25 PM »
Couple of things look mysterious on that drawing:
-4051 switches one input to eight outputs (or other way round) on your schem there´s just one signal in and one out? What do you want to achieve with that?
-4051 has three control inputs A,B,C. You can of course control them with a stream of gates, but it would only change one single bit from high to low, you need three streams to cover all eight switching options.

Please ignore me if I have misunderstood something vital, it happens ;D

Also the Lunetta section on electro-music.com houses some posters with very good knowledge
about "analog" digital CMOS, etc. in a synth context!
"Why not get an assistant to work the faders? Real vintage! And maybe cheaper... ;)"

substitute

Re: How to generate "words" for CMOS demux?
« Reply #8 on: June 03, 2020, 06:41:02 PM »
Quote
What do you want to achieve with that?

Always a good question. Specifically, I have a standard analog step sequencer with three parallel rows of eight steps of CV and a sequential switch module wherein you patch the gate of the last step into the control for the switch and the three rows of CV get switched making it a 24 step sequencer. The switch is activated at the end of each cycle, I'd like a device that counts cycles then switches. For instance four cycles of row1, two of row2, one of row3 then reset. That's the specific use case that got me reading datasheets, but I can think of other interesting uses to add variance and complexity to a simple sequence.


 

Related Topics

  Subject / Started by Replies Last post
11 Replies
3647 Views
Last post January 04, 2005, 03:29:01 AM
by Rochey
9 Replies
4736 Views
Last post February 27, 2006, 12:41:51 PM
by fucanay
16 Replies
2882 Views
Last post March 29, 2008, 03:16:49 AM
by Butterylicious
9 Replies
610 Views
Last post February 07, 2020, 11:06:58 AM
by Script