A few months ago I thought I found a schematic for a clock divider with three 4017 decade counters strung together to make a bunch of variable rhythms. I couldn’t find the schematic anywhere recently though so I tried to come up with my own. This is what I came up with. It seems to work all right. Any suggestions to make it better? I didn’t wire the switches up on the 1st 4017 the same as the other two because I thought having that 1/6 division would help make more interesting patterns.
1 Like
I tested it as a general frequency divider today. I just used my opamp LFO in high range and it sounds like it’s getting four octaves when switched up in such configuration. I need to get a better tuner than the one I have, but my ear doesn’t hate the result…lmmfao. I have a tuner on my peavy vyper but it doesn’t really do all that great all the time. 
1 Like
Kindly add a comment IN the pictures that the circuit is ‘UNTESTED’ or ‘TESTED’ or ‘WORKING’ so to prevent others from picking it up and not knowing what status the circuit is in. Images are gathered by search engines and any comments may get lost so if you put a comment IN the picture there is not risk of this happening.
One question. What is the reason for chaining the ICs via the transistors?
1 Like
Got it. I will update the pic when I get home. To answer your question (my best answer, anyway), I have been using that transistor setup to drive the led. I’m not educated in the mathematics or elec engineering of the circuit, but I read that it is a good way to buffer the output and drive the led simultaneously. Truthfully, I’ve not noticed a difference when just outputting off the 4017 pins directly, and having the transistor only for the led, but I’ve been told this circuit better preserves the signal strength.
It wouldn’t let me edit the picture again so I just posted a new one indicating that it has been tested and is working.
1 Like
You should probably buffer output out1, the way it is connected now, when the input clock is low, out1 is only pulled down through the 100K resistor in parallel with the LED and 1K.
The transistor configuration (emitter follower) that you use for buffering the other outputs is not ideal when you don’t know what will be plugged into it. This configuration will drive its output (emitter) high when its input (base) is high, but when the base is lower than the emitter, it does not drive it down, so again, the output is only pulled low through the LED circuit. So if e.g. you plugged the output of this circuit in something that has its input pulled high by default, your circuit might not work.
See for example this (admittedly a bit exaggerated) Falstad simulation with a 10K pulled high as the load circuit.
Regardless of whether you change the output buffering scheme, I would recommend connecting the second and third 4017 clock inputs directly to the output pin of the previous ICs, that’s how they’re meant to be connected and the transistor and resistor don’t help and what is plugged into the out could interfere with the operation of the circuit.
So should I just run all the outputs between ics directly from the ics pins, and add the 10k pull-up resistor at the output jacks from the emiters? And I should buffer output 1 this way as well?
Yes, one digital IC output can be connected directly to a few digital IC inputs of the same family. That is essentially what your are doing through the switches.
Not sure what you mean here, a pull-up would not help as the problem with the emitter follower is that it doesn’t pull down.
If you only want to generate 5V output pulses from your 12V supply, you could use the same output circuit as in your 4024 based circuit.
The main problem with that emitter follower based buffer is that it requires relatively small resistors which lead to relatively high power consumption. (On the order of 4mA when the output is high which might not be an issue as the LEDs already consume about 2mA each).
I haven’t done the math, but it might be worth it to use the usual opamp based buffer, if current consumption is a concern.
I’m sure I probably did not state that very clearly. LOL. I did however wire up the circuit as you suggested with the simulation and it seemed that the LEDs never completely turned off so I’m not sure if I’m doing it correctly. I am going to test the circuit I have with a few different modules that I’m going to build on breadboard. So far everything I’ve used it on has been built by me. That’s kind of the idea with this project is just to keep it working with things that I build. I may not end up changing it but I am not necessarily done experimenting with your suggestion yet. I did wire up the inputs on the ICS as you suggested. As well as adding the current buffer I’m using to the output one. I also changed all of the switch configurations to match the first IC with the 1/6 step. I’m planning to build tomorrow. 
I corrected an error on the schematic. The op-amp had a 57k resistor going to ground from the non-inverted input. It should be a 5k7.
Ah OK, my bad, I wasn’t clear.
The 10K in the Falstad simulation linked in my previous post was just a representation of a potential load from a module this output could be plugged into. It is not meant to be part of the output circuit, the Falstad simulation was to show how things could go wrong if such a load was driven by this output.
1 Like