Electronic wind chime

[RobertN]

Surely this has been done before, it appears to be something several PICAXE might handle.

The project can be more complex than it may appear, since from 1 to 6 tones can be sounding simultaneously, with each tone at a random exponentially decaying amplitude. The decay duration would be proportional a random strike amplitude. A light strike decay would be a couple seconds long, and a heavy strike decay greater than 10 seconds. Strikes can be interrupted with a new strike, which would increase the tone amplitude by the amplitude of the new strike. For a softer sounding strike, the tone should have a short constant rate ramp up to the random amplitude strike. Strike rate would be random with a possible bias to strikes occurring in clusters.

The initial thought is 6 08M's would handle one tone each, with an additional 08M selecting the tones in a quasi random manner. A real wind chime is arranged with complimentary tones adjacent to each other, and with tones being stuck with a bias towards adjacent tones. Low pass or band pass filters maybe required on each digital tone output for a more pure sound.

Does this sound feasible with 08M’s? How could the attack and decay of the tones be handled? Some of this may need an analog solution.

Any comments and suggestions appreciated.

 

[BeanieBots]

The overall spec sounds good and probably many 08Ms is the way to go. However, how would you tell a light strike from a heavy strike. QTC compound maybe?

 

[hippy]

Generating exact, pleasing tones will be another issue along with how to amplitude modulate the signal output. That will require some sort of analogue envelope to be generated and applied to the signal. Attack and decay is simply adjusting the analogue signal at the desired rate.

My opinion is that it's going to be a lot more complicated to achieve than it appears.

 

[Dippy]

Attack and decay can be done with op-amps.
But, as said, detection of force and control of envelope will be quite a task.

What is your level of electronic and code experience?
Have you had any experience of op-amps, vca etc. ?

If you are a novice, then this 'simple' project may get compilcated. It's a nice one though if you (and all those that help you on the Forum) have the time

 

[techElder]

I've been interested in these for a long time, but just haven't gotten going. I'm going to build a BIG one ... some day. It won't just depend on the wind for sound, either. It will have a keyboard with picaxes involved. You'll be able to play it.

You might join this group or visit some of the links. I found a ton of interesting facts that I never even thought of.

http://groups.yahoo.com/group/windchimeconstruction/

 

[Dippy]

You ought to build a really big one that has the complete string and percussion section of the LSO and generate some electricity too.

 

[gengis]

One alternative to the 'axe is to use single transistor oscillators. Years ago we did some twin tee oscillators with a variable resistor to set the feedback. Goose the base of the transistor with a pulse and it rings with an exponential decay. Low frequencies sound like tom-tom or undamped bass drums and higher freqs like bells.

Low parts count, and once "struck" it needs no further wave shaping and rings for a time.

The feedback is adjusted to a threshold where it won't support oscillation. One extreme of the pot and you have an oscillator, and the other a click; in between you can have a quickly damped bell tone or a more prolonged bell sound.

Actual bells are another alternative - simple push type solenoid striking a tubular bell or chime bar. It is something I'm interested in and working on these days. Why settle for the Westminster chimes when you could have a bar or two of "Pictures at an Exhibition?" Getting a solenoid to strike a bar isn't too hard - getting a nice quiet solenoid to strike a bar with a minimum of artificial scrapes clicks and clacks is more of a challenge.

Some on line research turned up one guy that does it using electric motors - turning only 180 degrees, to strike one of two chimes each actuator.

 

[techElder]

You could use a "motor" driven by an H-bridge to strike-hold-return at variable rates with an 08M at minimum.

 

[vttom]

What would be really cool is if you had some kind of sensor that measured the wind speed and you used that information to generate strike triggers sent to the other 08's.

 

[AndyGadget]

Some on line research turned up one guy that does it using electric motors - turning only 180 degrees, to strike one of two chimes each actuator.

I thought this was quite neat when I first saw it. Just the opposite of what the OP's after, but neat all the same, and a piece of cake with a Picaxe.
http://www.instructables.com/id/Arduino-Controlled-Bell-TowerCarillon/

Andy.

 

[goingsolo]

Take a look at http://hackaday.com/2010/04/14/electronic-wind-chime/.

It has fins attached to piezo elements to detect the wind. I would imagine that the stronger the wind the higher the voltage produced by the piezo.

 

[RobertN]

Some good ideas and approaches to electronic wind chimes were offered. Everything from purely electronic, to actual chimes struck in a random in a wind like manor. The electronic decaying oscillator is interesting, would have to read up on oscillators with exponential decay. A totally different approach would be to use the PicAxe tune command to excite an actual chime via a non contact exciter coil. It maybe possible to use the BPM and duration functions to vary the loudness of the chime by applying variable length tune signals in resonance with the chime. All look like a development project. This winter we may have to settle on a fan blowing on a chime.