frequency divider questions...

[11-90-an]

Ok, so i'm wondering what should be the value of R1 in this freqency divider schem....



(image is not mine, page 14 of engineer's mini notebook 555 timer circuits by Forrest Mims III)

This would be for a octave down that I'm trying to build with my current resources...

[Rob Strand]

Those things only work for a very narrow range of frequencies.   They don't really divide they just ignore pulses.

[ElectricDruid]

+1 what Rob said.

It needs a fixed frequency or it won't work reliably.

[11-90-an]

So if that's the case...

Is there a way to configure a 555 timer to be a frequency divider? When I google this I just find it as a pulse generator to be fed into a CD4017 for the dividing...

[Rob Strand]

Is there a way to configure a 555 timer to be a frequency divider? When I google this I just find it as a pulse generator to be fed into a CD4017 for the dividing...

I don't know of any way as the 555 only has set and reset inputs.  The 4013 can be used as a divider as it has a clock input,  It just happens to have set and reset as well.  Yes, 555 is normally a clock and the 4017 divides down the clock.  The 4017 also has sequenced outputs to do stuff.

[R.G.]

What Rob and Tom said. The 555 is a very flexible chip, and does have a flip flop inside it, but it is not practical as a wide range frequency divider. Any binary logic flip flop will naturally do frequency division by two when fed a pulse, and is a simpler thing to implement than the mess of stuff you'd have to put around a 555 to get it to work the same way.

In the general effects world, we naturally want chips that work at the nearly-standard 9V power supply naturally. The CD4000 series does this, so a CD4013 is, as noted, the natural choice. There are others, but this one is easy to work with.

A sneaky and somewhat deeper question is where are you getting the pulses that will run your divider? If you're hoping to run audio into the input of the frequency divider, you're going to be disappointed. Pulse based frequency dividers - and the 555! - depend on having a relatively clean input signal with nice sharp rise and fall slopes, no funny wiggling around in the middle. The process of cleaning up audio to make good driving pulses for a pulse based frequency divider is much more involved than actually making the pulse frequency divider.

[11-90-an]

I think first I would make a sine to square converter then feed it in...

I don't have any CMOS and I don't think I could buy some so that's whay i can't use 4017s or 4013...

[R.G.]

OK, you can do it a different way, with just resistors, capacitors, and NPN transistors.
Go read this:
https://www.electronics-tutorials.ws/waveforms/bistable.html
... especially the section titled Sequential Switching Bistable Multivibrator.

Using two transistors, you can make a binary divider.

[anotherjim]

If you were really pushed to use a monostable AND if you own an expression pedal, you could teach your foot to range the 555 monostable timing by adjusting a timing resistance with the pedal.
Typical pedal resistance is only 10k, although a volume pedal with 250k to 1M resistance might be to hand (look for 2 cable trick of using a volume/swell pedal as an expression). Using a 10k pedal to control the LED of a Vactrol with a 1M LDR dark resistance would be better.

A monostable divider can cover an octave, only the pulse width of the output gets too thin at the extremes. With sustained notes, changing the timing resistance will give a PWM effect. PWM is nice.

[ElectricDruid]

  I like this idea, Jim. It'd take a bit of learning, but it could be pretty good. Especially if you could organise the pedal to have a per-octave character, so that it wasn't all bunched up and super-sensitive at the low end. That's doable, even if it was only a log pot.

[Rob Strand]

If you were building a Boss OC2 style octaver, you could use a NE555 or 2xBJTs to replace the first Set-Reset 4013 then that would feed a BJT flip flop instead of the second D flip-flop 4013.



This ckt is similar to the Boss footswitch ckt.

[11-90-an]

If you were building a Boss OC2 style octaver, you could use a NE555 or 2xBJTs to replace the first Set-Reset 4013 then that would feed a BJT flip flop instead of the second D flip-flop 4013.

This forum really does contain mind-readers... 

I don't really understand what you mean by 2 BJTs replacing the first set-reset 4013... would the circuit end up looking like a flip-flop also? And would I also have to keep the lm324 SSC bit or just substitute this with a simpler SSC? 

[antonis]

Why you don't use a signal of 2 octaves lower and fully rectified it..??

[11-90-an]

Why you don't use a signal of 2 octaves lower and fully rectified it..??

Seems feasible... would be much noisier though... 

[Rob Strand]

I don't really understand what you mean by 2 BJTs replacing the first set-reset 4013... would the circuit end up looking like a flip-flop also?

The 2 BJT circuit I was referring to is the RS Latch (in older language you might see it called RS Flip-Flip) .  It's the circuit on this page.  Figures 3.22/3.23/3.24.   These are PNP versions but there are NPN versions.
http://www.tpub.com/neets/book9/36d.htm

The key feature of the RS-latch is it has two inputs: Set and Reset.   When you pulse Set the Q output goes high, a further Set pulse will not change the state.  When you pulse Reset the Q output goes low.   This is how the RS inputs of the CMOS 4013 work (more or less).  With the transistor circuit, the circuit is symmetrical so you just pick which output is going to be Q, then that determines which inputs are Set and Reset.

The circuit I posted above is a toggle Flip-Flip.  I only has a single input.  Each time you pulse the input the output toggles.

If you poke around the internet you will see a number of variations on these basic circuits, all very similar.

You might also notice some similarities between the RS-Flip- flop and T-Flip-flip.   They both have the two cross-connected BJTs.    In books and on the web you might find both of these called Bistable circuits.   Remember the number of inputs determines what circuit it is.  There's also two other similar looking circuits:  the Astable oscillator, and the Monostable pulse generator.

And would I also have to keep the lm324 SSC bit or just substitute this with a simpler SSC?

Keeping all that LM324 stuff certainly helps the tracking - that part *is* what makes it track.  The two circuits above just produce digital pulses.   Duck_arse was playing with simpler front ends to the toggle flip-flop and he found these didn't track so great; see his comment in the recent OC2 thread.  Some commercial octave circuits with simpler front ends are:
- Colorsound Octivider
- Maestro OB-2 Octave Box
- Shin EI OB2 Octave Box
Probably a few more.

The are all likely to track worse than the OC2.

[11-90-an]

Ok... could the 324 be replaced with other op amps like 2 tl072s? And could the JFET be repalced with a BJT? Just wanting to know before attempting and wasting time....

[antonis]

Just wanting to know before attempting and wasting time....

And of course, it's more convenient to waste our own time..!! 

[Rob Strand]

Ok... could the 324 be replaced with other op amps like 2 tl072s?

You can use TL072's but they are kind of wasted for circuits not passing the audio.   You can use LM358's which are a dual version of the LM324.  Some circuits used TL062's/TL064's to save power.

And could the JFET be repalced with a BJT?

That will cause problems you really need a JFET there.

[antonis]

And could the JFET be repalced with a BJT?

That will cause problems you really need a JFET there.

Did you take deep breath, Rob..??
('cause I'm pretty sure for OP's next question..)

[Rob Strand]

Did you take deep breath, Rob..??

Too tired to think - been up since 1:30am