frequency divider

[Pehr]

I'm planning to use a CD4040 as a frequency divider for a clock signal (0V/+5V square signal) to my sequencers, so that they can be run at different speed. That shouldn't be too complicated, right? Maybe used together with some inverters and some op amps it could work fine... Each stage divides the frequency by 2, and since it has 12 stages, I can have one sequencer go really fast and another one really slow, right?

But what if I want to divide a frequency by 3? or 5? 7? 11? 13? 15? How complicated is it to do that?

[zipzap]

I just finished a sequencer that does that. Gonna post some sounds in the next days. I use a 4017 for dividing by 1 to 10.
It´s really cool if you´re not fixed to ratios by 2. I did no buffering, since the divider is part of the sequencer logic. I just used the 0 output (pin3) of the divider 4017 as a clocksource of the sequencer. Outs 1-9 are sent to the reset in via a 10 way swich. The 10th position is left empty, the reset in has a pulldown to handle that. If the swich is set to connect 4017out 1 with 4017 reset that means divide by 1. The duty cycle becomes very long, your sequencer has to handle that. If it´s 4017 based it will. If not you´ll notice the sequencer stops when set to 1, while it can run at the other divisions. In that case use a 4538 monoflop to shorten the pulses.

[piedwagtail]

Hi,I find Ken Stone is the master of CMOS.

I've built his Pulse Divider with three 4017s and instead used a 4520 to do the 2,4,8.It does 2,3,4,5,6,7 & 8.As a fixed divisional system.I constructed on veroboard,with added 4093 clock oscillator and an added 4011 touch stop start switch.
There's no reason you couldn't incorporate further divisions with a 4040 or even /10 /11 /13 etc with strung 4017s.Chips wise it's easier to patch AND gates so for instance it only pulses when the 5 and 6 go high to give you /30.

With automation in mind his two votage controlled dividers are interesting devices.I've just completed both.
Jurgen Haible has a VCdivider too,but it was a little over complicated compared to Ken's designs.Both fire comparators to interrupt gates.

Here's the two boards for size comparisons....
Vero board left is post hack pre installation tidy....adding on 2 big 4067s to use the other side of the 4520.
Hence the long wires
Other chips Lm358,4011,4093 across the bottom.3 * 4017 and one 4520.Output transistors all 3904.


RS board on the right ,second generation VC divider above red long wire,first gen below.
Using a LM311 to fire out at the bottom.
I prefer to use thick Valve (6A 1000V) coloured single core so i can see what's going on.With the ultra slim wire that these boards were used with in the 70s,it looks beautiful but is harder to trace.

When i run out of these RS boards which were given to me (£25 a piece!) i intend to order the similiar Futuretech boards that Scott likes.Much cleaner finish than the spaghetti vero even counting the dozens of cut tracks underneath.I've had enough of vero...including the blisters from holing them!
Hope helps!
Robert

---

pict1099.jpg
Description:
Filesize:	154.55 KB
Viewed:	553 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[Pehr]

[bugbrand]

I'd def. agree with the goodness of those Ken Stone circuits - I've done the Master Divider and the Pulse Divider - actually, note that there's an ooops-mistake on the PulseDivider Scheme at the moment - on the 2nd 4017 the reset should be from Q8 (pin9) not Q9 (pin11) - Ken says its corrected on the PCB but the schem error got me very confused.!.

The non-2-4-8 (non-binary?) divisions are very very great indeed 'cos you can have regular rhythms running but have wierdy cross-rhythms over the top - good results aplenty!

[Uncle Krunkus]

You can also just AND the original clock with the pin3 output to retain the same pulse length.

[zipzap]

Are you sure? Similar things didn´t work for me. The delay of the 4017 makes he clock come in too early (i´d rather say: The 4017 is too late).
Can fix it with one or two aditional gates, of course.

[Uncle Krunkus]

I have only tried it once. I didn't notice any appreciable delay, (the 4017 has switch delays in the nsecs doesn't it?)