BS2 & Piezo Speaker

Humanoido

Looking for some VERY SIMPLE ways to increase the
volume of a piezo speaker on a BS2 HomeWork
Board...

PJAllen

"...VERY SIMPLE..."
How about a paper cone, the old Victrola effect?

Post Edit -- Works great with a funnel.

Post Edited (PJ Allen) : 11/1/2008 5:40:16 PM GMT

Mikerocontroller

· humanoido, have you seen the resonator used on the SoundPal?· Maybe a section of thick shrink tubing?

Carl Hayes

A simple cone would be surprisingly good. Even better would be a cone shaped like the bell of a trumpet, but of course smaller. With the right shape you can transfer much more energy into the air from that tiny transducer. Edison mechanical phonographs had diaphragms not much larger than the one in the little piezo speakers, yet with the big flower-shaped cones they were easy to hear. Unless, of course, you're ancient like me, and have trouble detecting freight trains ten feet away.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i

Carl Hayes

Come to think, you could get a tiny plastic toy instrument and cut off the cone.· I may try that myself.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i

Humanoido

That was in the back of my mind about a tiny cone.
Per suggestion, I tried it. Took a yellow 3M Post-It Note,
folded into a cone shape and slipped it over the piezo speaker.
Gadzukes!!! It appears to have doubled or tripled the volume!!!

Another great idea about using the heat shrink tubing...
I looked through about 20 different sizes in the storage bin,
but none were large enough to go over the piezo speaker's
diameter. Then I remembered, I bought a Chinese washer
drain hose to use as strain relief for robot servo cables.

I cut off a small cylinder piece and place it over the piezo.
There were two pieces. One from the end which had smooth
walls, and one from the corrugated section. I expected the
smooth would work better.

The results are in. The smooth and corrugated sections work
the same with no noticeable difference. Comparing the
cylinders to the cone, performance is nearly the same, with
the cone taking 1st place for best overall frequency range
resonance. The tiny cone can reproduce lower frequency
with richer sounds.

It's much more convenient to slip on little cut pieces of
cylinder than to fold up paper cones. The frequency response
favors the upper end sound spectrum where higher pitches
have excellent volume.

My next question - does the length of the cylinder or size of
the cone matter? If the cone or cylinder is made larger and
larger, at what point is the sound maxed out?

Any guesses??? [noparse]:)[/noparse]

SRLM

A 5000 Hz sound wave has a wavelength of about 2.7 in, so that's probably why the corrugated and smooth worked the same.


For making it longer, you'll probably have to guess. I'd assume that each bounce of the wave on the wall takes out a little bit more energy from the wave, until it just dies. If I had to pick a number, I'd say a length of 1.35 in would be ideal, but I don't really have any idea at all.

MovieMaker

do you have any pix? I would like to see, if it is possible.

Carl Hayes

Ideally the sound waves don't bounce off the walls of the cone anyway.

For an excellent treatment of cones, or horns,·used for coupling speakers to air, go to the library and see if they have the·Radiotron Designer's Handbook, Langford-Smith,·circa 1953.· This was published in the USA by Radio Corporation of America.· It's a very fat (3 in)·red book with an incredible wealth of information.· Start on page 851, "Horn Loudspeakers".

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i

Humanoido

Here's the pic showing an experiment with
a half side open directional speaker cone.
In the background you can see the tubing
which also works well.

erco

Transistor amp. The BS2 HW board has 220 ohm resistors inline on all the IO pins, which:

a) limit how much current can flow (bad), and
b) are perfect to bias a 2N2222 transistor amplifier (good)

I know this provides a dramatic volume increase on a paper speaker (mit kapazitor, jah?) and it should help a piezo speaker, too.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Humanoido

erco, got a simple schematic? I'll build it.
Should have some 2n2222's just
waiting to try it out.

Beau Schwabe

This topic comes up every now and then... "I want a louder piezo" ...If you understand what the piezo crystal is doing and how it behaves, there are certain things you can do to exploit the behavior of a piezo.

Reference page:
http://forums.parallax.com/showthread.php?p=464410

Low power Piezo inverter driver schematic:
http://forums.parallax.com/attachment.php?attachmentid=36227

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

erco

Looks like Beau beat me in timeliness and technology! His piezo solution is definitely·better, driving the piezo crystal·in both directions. My simple approach is below, just using one transistor. It is probably more effective on a traditional paper cone speaker since it lets more current flow, but it may yield improvement on a piezo speaker as well. Keep in mind, my drawing shows a 220 ohm resistor (built into each of the HW board's 16 pins) so don't add that to a HW board. Only add it to a discrete Stamp or BOE.

Using the same schematic below,·you can·also attach your piezo's positive lead to +9 volts instead of +5 volts for more voltage & volume.

BTW, the resonant frequency of your speaker is a huge factor, per Beau's excellent links.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Tracy Allen

Another simple solution is an audio line transformer. I have the Xicon ultra-mini, 600 ohm to 100 ohm, with the 100 ohm side to the Stamp or sx or prop, and the 600 ohm side to the piezo. (Mouser 42TL031-RC) It can be used in combination with push-pull drive that Beau suggested. Note that if you are doing this with an sx chip or a prop, you can get the push pull drive by connecting between two output pins, driven 180 out of phase.

I use this with ultrasonic piezos. The transformer is not rated for 40khz, but it does in fact work happily there.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com