Maker Pro
Maker Pro

Looking for oscillator and frequency divider schematics

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
I'm looking for the schematics for a square wave generator for an electronic organ.
Ideally, I'd like to be:

- as simple as possible
- only use discrete components, no IC
- be reasonably stable frequency-wise
- have a 50% duty cycle
- simple to adjust the frequency, preferably by changing the value of only one component
- as a bonus extra, be able to vary the frequency slightly with a control voltage

I've been looking around, and astable multivibrators, but adjusting the frequency seems to involve two capacitors and/or two resistors. The other thing I thought of is using a phase shift oscillator followed by an inverter to turn the sine into a square wave, but it semms a rather crude solution.

I'm also looking for the schematics to a frequency divider. I know you can use a flip flop to halve the frequency, but I have only seen D flip flop used this way. Is there a simple circuit that can easily be built with tiscrete components to do tha same thing?

Can anyone help?

Thanks in advance
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
I think a Schmitt trigger oscillator built with discrete components would work. The feedback resistor would set the frequency. You could even probably use a control voltage through another resistor to vary it.

Edited: Welcome to the forum!

Bob
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
As simple as possible does not agree with discrete components.

The simplest oscillator that I know is a Schmitt trigger with a capacitor from input to ground and a resistor from output to input. The output will be square wave. A 4093 will give four oscillators and a 40106, six.
I do not know the frequency constancy but should not be too bad with a stabilised supply.

The 4024 divider will give seven outputs so you get eight octaves.

Modern organs use computers to generate the sounds, this gives much variability, it is all done by software.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
An astable multivibrator is a simple discrete solution. Use a 2 gang pot to adjust two resistances at once (it's much harder to adjust 2 capacitances -- especially of the size used in this sort of circuit)
 

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
I somehow managed to miss the SChmitt trigger oscillator in my searches, it looks exactly like what I need.

Any ideas about the circuit to halve the frequency?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
You could use a T flip-flop
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Somewhere I have the circuit for a Livingston Monarch organ. I could look this up if you are interested. It uses germanium transistors.
The top octave uses a pot core oscillator and the dividers are discrete components.

Using dividers means that the octave frequencies are locked to each other. You may get a more interesting note if you use a free phase organ with non locked harmonics. This would entail 60 oscillators for a 60 note keyboard.

I gave the organ away as my playing skills were non existant.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
As duke37 already suggested, you can use a CD4024 to generate seven lower octaves from each oscillator. (I'm assuming you will have twelve oscillators, one for each note, so you will need twelve CD4024s.)

If you really mean discrete, i.e. no ICs, look at the schematic for the transistorised clock on kabtronics.com. It uses two-transistor flip-flop circuits to divide frequencies by 2. As you can see, though, it takes a lot of space!
 

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
Thanks for the replies!
I think I've found what I'm looking for, once I've sorted a few things out I might start building a prototype.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Here is the heart of the Livingston Monarch. It uses pnp germanium alloy junction transistors and runs on -12V. Now, you should use npn transistors run on less than 6V, because of the limited base/emitter voltage.

You will need a LOT of components to get a wide range instrument.
 

Attachments

  • monarch.png
    monarch.png
    58.7 KB · Views: 231

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Thanks for the replies!
I think I've found what I'm looking for, once I've sorted a few things out I might start building a prototype.
Keep us posted. I would like to see what you come up with.

Bob
 

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
Well, here is what I've found:

This is the oscillator: http://www.bowdenshobbycircuits.info/page13.htm#schmitt.gif
And this is the frequency divider: http://www.electronics-tutorials.ws/waveforms/bistable.html (near the bottom, the squential switching bistable multivibrator)

I'm just getting my head around these circuits, being a bit of a beginner at this stuff, and I have a couple of questions:
In the oscillator, the value of R3 depends on the current drawn by the load, which in this case is a chain of frequency dividers. Do you think this value is critical, as long as its not too high? If it is, how can I calculate the current drawn?

Thanks again, I'll keep you posted.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
The value of R3 isn't critical. I would choose a value that gives a current of around 1~10 mA, unless there's some reason not to. So if your power supply is 9V, a value between about 1k and 10k will be fine. For example, 3k3.

The frequency divider circuit doesn't draw much current. And it's only the first divider that you're driving from the oscillator anyway.

That oscillator will not produce a square wave with a duty cycle of exactly 50%. You probably want a 50% duty cycle to ensure the right mixture of harmonics. You can solve that problem by running the oscillator at twice the highest frequency you want, and taking your first output from the output of the first frequency divider. (The output from a frequency divider will always have a 50% duty cycle.)

For the frequency divider, I'm assuming you intend to use the "sequential switching bistable multivibrator" circuit. This should work fine, but there are quite a few components in it, which is significant if you plan to build lots of them. Did you check out the kabtronics design? It's similar, but it might be a bit simpler.
 

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
Why will the duty cycle not be 50%? Is it because the value of R3 is assumed to be small when it comes to choosing the values of R2, R4 and R5, that set the switching thresholds?

Is the kabtronics one the circuit on page 33 of the transistor clock manual? It is simpler, I think I'll use it rather than the other one I posted above.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
I would not worry too much about the variation from 50% duty cycle. A 50% square wave has only odd harmonics and a sawtooth wave is preferred which has even and odd harmonics. Hammond made an organ with square wave generators but modified this by adding harmonics to give some even harmonics, this gives a richer tone.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
The reason why the transistor Schmitt oscillator's output won't be 50% is that the switching thresholds are not exactly the same distance away from the supply rails. So the capacitor will take longer to charge to one threshold than it takes to discharge to the other threshold.

It's actually quite difficult to generate a square wave with a duty cycle of exactly 50%. The best way to do it is using a divide-by-two stage, as I suggested. If the duty cycle is not exactly 50%, the frequency spectrum will include unrelated frequencies as well as the fundamental and its harmonics.

Yes, the kabtronics manual page 33 is the circuit I was recommending, but looking at it, I think the one you linked to is a lot better. I'm not happy with kabtronics's explanation of the trigger/toggle behaviour, and I think the design you linked to, with the 220k resistors, is better. The 100k base-emitter resistors shouldn't be needed though.

Also the component values given are for relatively slow operation; for operation at audio frequencies I would reduce the 220k resistors to 100k and reduce the 1 nF trigger coupling capacitors to 100 pF or so.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
In the circuit posted, there is a suggestion on how to get a 50% duty cycle by bypassing the feedback reistor wtih another resistor and diode.

Bob
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Bob, thanks, I saw that. It won't be very accurate. For musical applications I think it's important to avoid any imperfections in the frequency spectrum. OTOH, if the signals will be heavily filtered, it won't matter!
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
And your extra flip-flop is the gold standard solutuion.

Bob
 

Erudhalion

Mar 25, 2013
12
Joined
Mar 25, 2013
Messages
12
Ah, I misread the part about the 50% duty cycle, I thought the extra resistor was to get a duty cycle other than 50%. I'll see how far from 50% it is, and decide what to do then.

EDIT: And another thing: what supply voltage do you thing the frequency doubler is meant to run at? I can't semm to find it on the site.
 
Last edited:
Top