I am bending a little toy and when shaking I will make it glitch the pitch. A glitch maracas.

It has a small container inside perfect for making some tilt switch, but I have not succeeded making it work satisfying.

I have among other things tried to put aluminum foil inside, four touch points, bolts and screws, but it does not trigger well.

So do you have any experience with this and if so how should I do it? I have seen that there is a solution with a "boxed mercury ball".

Last edited by Dip (Jul 5, 2016 11:16 pm)

A very long time ago I got some tilt switches. They consisted of a plastic box with a ball bearing (about 1/4") and 4 solid wire contacts on the inside walls. The box is about 1/2" x 3/4", so the ball can only travel about 3/16" along the long axis and only 1/16" between the short axis walls.
  The way the wires are setup- 2 in the center on the bottom, the ball rests on these when the box is level. Then a contact at either end of the long axis about half ways up; so when the box tilts, the ball bridges one of the bottom and a side contact. And when the box levels again the ball goes back to center on the 2 bottom contacts.
   So these are only a single axis and not very precise. They could be wired to detect level or a left or right tilt. Hope this makes sense.
Yogi

....

Last edited by Domu (Jul 5, 2016 8:20 am)

A piece of wood with a circle of tiny nails holding a metal ball bearing has worked for me, but not sure you could do that inside your toy by the sound of it... Good luck

accelerometer?

yogi wrote:

A very long time ago I got some tilt switches. They consisted of a plastic box with a ball bearing (about 1/4") and 4 solid wire contacts on the inside walls. The box is about 1/2" x 3/4", so the ball can only travel about 3/16" along the long axis and only 1/16" between the short axis walls.
  The way the wires are setup- 2 in the center on the bottom, the ball rests on these when the box is level. Then a contact at either end of the long axis about half ways up; so when the box tilts, the ball bridges one of the bottom and a side contact. And when the box levels again the ball goes back to center on the 2 bottom contacts.
   So these are only a single axis and not very precise. They could be wired to detect level or a left or right tilt. Hope this makes sense.
Yogi

That makes sense, thank you. That is pretty much as I had it temporally set up but I made the box and a thing inside myself, and it was barely triggering.

@Domu
Yes, that is a problem and it is hard to avoid, its just a matter on how much noise is acceptable. Or even valuable.

@SketchMan3
I never even thought that an accelerometer would be possible in a simple bend. I figure it needs power and connection to a computer? Pricey? If it would be doable it is awesome.

By shaking the tilt box you only get a short signal at a fixed level (since the ball bounce away), and that is not so good (it's a tilt box, not a shake box). Ultimately you should be able to shake and affect the sound by shaking harder (or possible by changing axis a bit).

One idea is to on the inside make a (pencil) graphite ribbon resistor and have a ball or something hit that. Then the glitch pitch would be changeable.

It could would work with a small nail bent around another nail and let it tilt between two touch points, for just a simple trig. Problem is that it will (also) bounce off. Maybe with a weight at the end of the nail and a soft rubber band holding it back would be better. Still only fixed pitch.

Maybe put in photoresistors to detect motion (but accelerometers would be better I guess). I will probably put in one or two photoresistors anyway, for other effects.

big_smile

THis is a U.S. Company, but if you have an electronics overstock/junker over there, these are usually cheap:
http://www.goldmine-elec-products.com/p … ber=G21036

They are very commonly used in assembly lines, and still manufactured in various sizes.

THis is a U.S. Company, but if you have an electronics overstock/junker over there, these are usually cheap:
http://www.goldmine-elec-products.com/p … ber=G21036

They are very commonly used in assembly lines, and still manufactured in various sizes.

Dip wrote:

@SketchMan3
I never even thought that an accelerometer would be possible in a simple bend. I figure it needs power and connection to a computer? Pricey? If it would be doable it is awesome.
big_smile

I honestly have no idea. Just throwing out an idea that i heard Harmonix was using for their guitar controllers now, lol.

What's the inside of your toy look like? I'm thinking a contained peizo contact with something to bounce against it. Beans, sand, idk.

Also, what kind of range are you looking or? Do you want a redo ate trigger each time you shake, or did you want a range of resistance depending on how hard you shake?

Last edited by Jazzmarazz (Jul 7, 2016 3:28 pm)

Thanks for all the input!

@vblank

I found one (tilt switch)! Very cool.

@SketchMan3

Ok, cool. Ideas are good.

@Jazzmarazz

The space inside the toy (cellphone) that I plan to use is something like 2x3x4 cm.

[The space is right behind the speaker and the magnetism from the latter is strong enough to affect objects in the little space... But maybe that actually can be a good thing]

The piezo sounds like a good idea, and I happened to buy a bag of them the other day. Ultimately the resistance would be depending on how hard you shake, yes. The range can be as large as possible, I think the pitch should be higher in the accent, louder, position.

-----

Possible ways(?):

[I have discovered these glitches: pitch up/down (wire + or - to clock), gain/distort, and mute (of course). Also some soft gain that change the overtones.]

a) A tilt switch that bypass a gain boost or mute the signal completely. Loud in default position, soft when moved away.
Plus a fotoresistor/LDR (I got one of each, do not know which one is better to use) in the antenna. Then the pitch will also change when shaken, I suppose.

b) a piezo to clock faster/pitch upwards (or maybe one in each direction?!). The loud gain glitch only works with "full bypass" of the circuit points, so I suspect a piezo will not work with that. But bypassing one of the speaker wires might work.

These ideas sound pretty good to me, clean and simple enough! B)

EDIT: I found this (old) mouse ball which could hav been perfect for the piezo variant, heavy and with a rubber coating (silentish shaking), but it is too big. May be of use in another project.

Last edited by Dip (Jul 8, 2016 12:03 pm)

I made something similar using a cheap pedometer. The pedometer had a small lever inside with a weight at the end attached to a loose spring so that it would connect the circuit momentarily each time you took a step (or shook it) it is a simple on/off mechanism though so if you are looking for variable resistance for a pitch bend I'm not sure it would work.

Last edited by SamVsSound (Mar 1, 2017 1:08 am)