First attempt at a 555 timer "instrument"

[kalel]

This is loosely based on this circuit:
http://www.555-timer-circuits.com/toy-organ.html

By loosely, I mostly refer to making a mistake in design and also not using the correct components, just what I had to experiment with.
That means, most of the notes are off and none is really in tune.

I never used a PCB design tool, and then I rushed this design to take advantage of a coupon, as the PCB cost me $2 shipped. I didn't connect everything well, costing me one button and resistor (it could probably be fixed with some wires but is not that important).
With that in mind, I'm amazed that it even works at all. With the right component values, even the tune might be slightly better.

So, the result is not great at all, but it works!

(the sticky stuff is mostly glue, and not flux, to try to insulate things from fingers).

Here's a short snippet of it recorded using a microphone, then filtering with a tool called Ignite. The filtering is mostly adding some reverb, the out of tune sound is intact.

http://picosong.com/wsymx/
(If you do have a listen, don't expect an enjoyable experience, only 4 buttons are slightly in tune relative to each other, others are mostly useless)

As a potential redesign, perhaps keeping a piece of the board that connects the 555 timer itself, and then having each note have its own trimmer pot and button might be a good choice. They can be added on a protoboard which would connect to the main board to shape the resistance. That would allow independent tuning of each tone. There's also the fact that these things are much more easier to do with something like Arduinos. There's also the fact that using these PCB push buttons, even if everything was in perfect tune, playing would not be easy. But it's good to keep in mind that this circuit is just for fun, and not any actual playing.

[Alex Eisenhut]

https://en.wikipedia.org/wiki/Stylophone

[kalel]

https://en.wikipedia.org/wiki/Stylophone

Great thing, but what generates the sound?

[Ian.M]

http://hackaday.com/2012/07/22/reverse-engineering-a-stylophone/

[Alex Eisenhut]

https://en.wikipedia.org/wiki/Stylophone

Great thing, but what generates the sound?

The speaker! All kidding aside, the schematics I've found for the original show a component that isn't even made anymore; the unijunction transistor. This makes it a relaxation oscillator if I'm not mistaken.

This thing originally came out before the 555.

When I was a kid this song came out



I had no idea what a stylophone was, I thought they  meant the stylus of a record player and a stylophone was some sort of record player.

[Alex Eisenhut]

Huh, they still make "programmable" UJTs, look at that!

https://www.digikey.ca/products/en/discrete-semiconductor-products/transistors-programmable-unijunction/290?k=ujt

They even still make these:

http://www.newark.com/multicomp/2n4870/unijunction-transistor-0-3w-50ma/dp/35C0717

Wow. I guess I should have said "I haven't seen or heard from them for over 20 years." 

[kalel]

Huh, they still make "programmable" UJTs, look at that!

https://www.digikey.ca/products/en/discrete-semiconductor-products/transistors-programmable-unijunction/290?k=ujt

Some metal cans seem to be available from China.

[Zero999]

One of the disadvantages with the 555 is it can only make one note at a time. You could use two 74HC14s to make 12 astable multi-vibrators, each tuned to a different note and connected to a mixer circuit to enable chords. You may also find the sawtooth waveform, taken from the capacitor sounds nicer than a square wave, but you need a high impedance buffer, to avoid interfering with the circuit.

[Alex Eisenhut]

If you want I can mail you my 2n4870s, at least you know they're genuine.
I last used these in school 20+ years ago and forgot about them, but they're still in my parts cabinet...

PM me your address I'll see if it's cheap to ship.

How's the postal service in Afghanistan these days? Reliable?

[kalel]

If you want I can mail you my 2n4870s, at least you know they're genuine.
I last used these in school 20+ years ago and forgot about them, but they're still in my parts cabinet...

PM me your address I'll see if it's cheap to ship.

How's the postal service in Afghanistan these days? Reliable?

That's kind of you, but my knowledge is not much beyond using 555 so far. I've yet to try the other oscillator circuits before attempting a more complex solution with them.
Other than that, a good stylophone also needs a good case and PCB design most likely (so you can nicely "slide" over the tones). The pen can probably be improvised from something, it should be smooth enough not to cause damage.

So for me, I think a Stylophone project is out of my reach for now (could get a bit costly to do it right too). It is interesting to know at least the basics principle it works on. I remember seeing videos about it.

[kalel]

One of the disadvantages with the 555 is it can only make one note at a time. You could use two 74HC14s to make 12 astable multi-vibrators, each tuned to a different note and connected to a mixer circuit to enable chords. You may also find the sawtooth waveform, taken from the capacitor sounds nicer than a square wave, but you need a high impedance buffer, to avoid interfering with the circuit.

Well, the only solution would be to use multiple 555s. But for polyphony, it should also have at least some more keys. Sure, you could do some chords with 7 tones, but it won't make a huge difference.
I wouldn't know how to get proper keys either for a more complete solution (these PCB ones are not easy to play with, perhaps some other types that come with caps), alternative is perhaps getting a toy instrument or some old malfunctioning keyboard that still has the keys intact, and then replacing the circuitry (with the far inferior solution to whatever is in there ).

[Ian.M]

I don't see the problem with case or PCB design for a Stylophone for geeks - it would be entirely usable as a single double sided bare board with a rubber foot at each corner, with the stylus pad along one edge with a mask cut out from foam board or 3D printed to provide a perimeter for it.  There's nothing special about the stylus pad - ENIG would do at a pinch though HASL would probably be more durable.  The premium option would be electroplated hard gold - but you'd have to be careful with the board layout to get the entire stylus pad within the required distance of the edge for your PCB manufacturer's edge connector plating bath.

[orolo]

If you want I can mail you my 2n4870s, at least you know they're genuine.
I last used these in school 20+ years ago and forgot about them, but they're still in my parts cabinet...

PM me your address I'll see if it's cheap to ship.

How's the postal service in Afghanistan these days? Reliable?

I have a bunch of vintage 2N2646 UJTs in metal TO-18 package, and no use for them. As Alex said, I'm not sure about shipping costs or reliability, but PM me if interested.

[kalel]

Thanks everyone. I don't think I will be making a stylophone (at least yet, perhaps in the future?) but your help is appreciated. I'm partly to blame for heading the topic towards it as I was interested in how it was made.
I guess that due to the construction it also has monophonic sound (one pen, one contact).

With some effects, it sounds pretty cool:


Those effect pedals might not work so well on 555 timer based things, but I'm not sure (you would have to try). Although at some point it gets better to use a full blown keyboard (or a MIDI one for the computer) if you want more sound effects.

It might also be interesting if there was a variable resistance model where you could slide between notes. More difficulty, but more capability.

[Ian.M]

There's no need for difficult to obtain parts as a Programmable Unijunction Transistor is a simple four layer PNPN device similar to a SCR (Thyristor), and an equivalent circuit element can be assembled from a complimentary pair of 'jellybean' transistors both connected with their base to the collector of the opposite one.   The PNP emitter becomes the Anode, the NPN emitter the Cathode, and the PUT gate (anode gate) is the PNP base.   (If you want SCR operation, use the NPN base as the cathode gate.)



LTspice sim attached.

The only significant constraints are that the bias point for the gate shouldn't exceed the PNP Vbe reverse breakdown voltage, and it isn't quite as robust as a real PUT as the transistor bases each have to carry about half of the peak discharge current.   Lower gain transistors are often preferable.   Its essential that both transistors' max Vceo  is several times the gate bias point as the Iceo leakage current mst be negligible.

@Kalel: You should  have suitable parts in your junk box - give it a try.  Just use any NPN and PNP small signal transistors with Vceo_max >=30V and Ic_max>=600mA.  RV1 is where the Stylophone stylus pad resistor chain connects.

[Zero999]

One of the disadvantages with the 555 is it can only make one note at a time. You could use two 74HC14s to make 12 astable multi-vibrators, each tuned to a different note and connected to a mixer circuit to enable chords. You may also find the sawtooth waveform, taken from the capacitor sounds nicer than a square wave, but you need a high impedance buffer, to avoid interfering with the circuit.

Well, the only solution would be to use multiple 555s. But for polyphony, it should also have at least some more keys. Sure, you could do some chords with 7 tones, but it won't make a huge difference.

That's the beauty of using the 74HC14. It contains six Schmitt triggers, so only two ICs are required to make 12 notes, which is enough for most tunes. The only disadvantage is the frequency is less predictable than the 555, but once you built a circuit with a certain IC, measured the frequency and compared it to the calculated frequency, the scaling factor can be applied to the other oscillators with that IC.