Reinhard's Experimental Physics Letters (unpublished) 5/1996
Single Bubble Sonoluminescence HOWTO
What's this all about?
There are many papers about the theory of Single Bubble Sonoluminescence
available, but exact descriptions how to produce it are rare. So
if you already know about sonoluminescence and now want to reproduce it,
this is the right place to look at. I tried to give a complete and detailed
report of the steps towards SBSL. Any suggestions, supplementations, comments
This experiment may be dangerous. Always think before you work.
Working with vacuum or boiling water can cause explosions.
High voltages are generated in this experiment.
I will not be responsible for any injuries or damaged equipment.
If you don't know what you are doing just don't do it at all.
And: No, I am also not responsible for any
thermonuclear accident you trigger.
- sine generator:
any function generator working around 25 kHz, adjustable to +/- 1 Hz
(+/- 10 Hz may work, too)
nearly any kind of audio amplifier will do. If you're not sure, measure
the saturation voltage: 40 V peak-to-peak should be enough.
- 2-trace oscilloscope
- 2 piezoceramic Transducers (drivers):
around d=16 mm in diameter, h=8 mm thick
- piezoceramic pill-transducer (microphone):
around 3 mm in diameter, 1 mm thick
- three finger clamp
- laboratory stand
take a 100 ml Pyrex/Duran spherical flask, diameter 65 mm, with
a small neck. An industrial one has poor optical quality, so better take
a free blown one.
- coil(s): around 20 mH, see text
- resistors: 1M, 10k, 1R
- coaxial cable
- quick-drying epoxy glue
- an eyedropper or a syringe (one of those little do-it-yourself subcutaneous
is very good)
- degassed distilled water:
- Pyrex/Duran Erlenmeyer flask (0.5 or 1 l) and airtight stopper
with pipe, rubber hose and clamp to close it
- aluminium/highgrade steel drinking bottle (0.5 or 1 l) with screw
cap; one of those found in camping stores, a bare one without varnish
- a bubble ;-)
Equipment (nice to have):
- second oscilloscope
- vacuum pump
- high-pass filter
Set up the assembly:
- Prepare the transducers:
They are polarized, so when connecting them in parallel watch out for the
mark on the electrodes. Do not connect them antiparallel. Clean
the electrodes of the transducer with a pencil eraser. Do not erase
the mark. Put 3 dots of solder equidistantly on one electrode; This will
give a better contact when glued to the flask. You may also connect 3 wires
to the dots to have spares in case one breaks. Put one dot on the other
electrode. Be careful soldering on the piezoceramic: Use a cold (250 degrees
centigrade) soldering iron and work quickly to avoid reaching the
Curie Temperature (see transducer datasheet; 320 degrees centigrade
for PIC 155, but most of them got an even lower one). Use fine wires
to reduce sound loss. Tin the wires and solder them on the dots. In the
same way put a wire on each electrode of the pill-transducer. These wires
should be really short before entering a coaxial cable. It's also
possible to connect a thin coaxial cable directly to the pill-transducer.
- Glue the transducers to the flask:
Clean the glass with aceton. Fix the drivers on opposite sides of the flask.
Remember the marks on both transducers have to point into the same direction
i.e. both towards the flask or both away from it. Use a quick-drying epoxy
but watch out: Some of them shrink when they set, causing the glass to
crack. If this happens:
- take another epoxy
- take less epoxy --- cover the whole surface of the driver, but not
- first attach the spherical form to the drivers, e.g. by using a separator
between flask and epoxy. Between the resulting fit and the flask only a
thin layer of epoxy glue is required, which will cause no trouble.
- take a lower diameter of the drivers
Fix the microphone transducer on the bottom of the flask.
Attach the flask with the three finger clamp to the laboratory stand.
select the coil(s):
One final hint: You really have to glue the transducers. Don't try
any clamp-press-or-something-stuff. It won't work.
Measure the total capacity C of your drivers. A transducer makes quite
a nice plate capacitor, so the following calculation may give you a hint,
C=13.9 e (d^2)/h [pF/m]
where d is the diameter, h is the thickness and e is the dielectric constant
of a transducer (see datasheet of your transducers, 1700 for PIC 155).
To match this capacity to the audio amplifier, you have to set up a serial-resonant
circuit with a coil of the inductance:
L=1/(C (6.28 f)^2)
where f is the resonance frequency (about 26000 Hz for the used flask).
Example: Two PIC 155 transducers with d=16 mm, h=8 mm, e= 1700
have a calculated capacity of C=756 pF. At a driving frequency of
f=26 kHz the needed inductance is about L=50 mH.
The inductance has to be variable. This is achieved by an adjustable core
or by changing the distance between two (or more) coils wired in series.
Last not least the coil has to be strong enough so it won't overheat.
Connect the generator to the amplifier. Connect the coil in series
to the driving transducers and the 1 Ohm resistor (current sense).
The other two resistors make a 1:100 voltage divider at the input of the
resonant circuit (voltage sense). Use coaxial cable where ever possible.
Fix the cables to the laboratory stand to avoid wire breakage. The voltage
of the resonant circuit may give you a shock, so don't touch it, or even
much better: Insulate any exposed connection with suitable tape or varnish,
especially if you want to produce Multi Bubble Sonoluminescence.
Persuade one bubble to glow:
- Prepare degassed Water:
Clean the flask thoroughly.
Open your vessel with the degassed water. Pour the water into your
flask letting it run down the glass. Be careful not to produce bubbles
while filling. Fill it up to the neck, so the water is nearly spherical
Find the acoustic resonance:
- with a vacuum pump:
Fill the Erlenmeyer flask or the bottle up to the half with distilled water
and pump off the air until the gas is completely pulled off the water (at
least 15 minutes).
- with an Erlenmeyer flask:
Fill the Erlenmeyer flask up to the half with distilled water, close it
with the airtight stopper and heat it. Leave it boiling with an open hose
for about 15 minutes. Take it off the heat and clamp the hose. Let it cool
- with a drinking bottle:
Fill the bottle up to the half with distilled water, put the cap loosely
on top allowing the air to escape and heat it. Leave it boiling for about
15 minutes. Take it off the heat and close the cap completely. Let it cool
down. Don't bother me with "my bottle exploded, what was wrong?"-stuff;
While the bottle is heated it mustn't be closed!
By-pass the coil and set the sine generator to 26 kHz. Display the
microphone signal on your oscilloscope. Now change the frequency slowly
to find the maximum amplitude of the signal. There may be several local
maximums around that frequency, but in most cases you'll have to select
the absolute maximum (i.e. the highest peak). If you got a sweep generator,
it might be helpful to display a 20-30 kHz
sweep on the scope. The acoustic resonance appears as a some 100 Hz
broad peak. If you found the resonance, remove the by-pass.
Adjust the inductance:
Display voltage and current on your oscilloscope so you can see the phase
shift. Adjust the core/change the distance between the coils until the
phase shift is zero. If you got a sweep generator, it might be helpful
to display a 20-30 kHz
sweep of the microphone signal on the scope. The electric resonance
appears as a 1 kHz broad peak overlaying the acoustic resonance.
Choose a low driving amplitude, e.g. 100 mV peak-to-peak on your
Watch the microphone signal: It shows a pure sinusoidal voltage. If
ripples are on the signal, there are some bubbles in the flask. Leave them
alone a minute or two with the driving voltage switched off. If they don't
vanish, your water contains too much air; you have to degas
it more carefully. Ripples may also indicate a poor contact between
a transducer and the flask. The epoxy has to cover the whole transducer
and no air may be between transducer and flask. Air between epoxy and flask
appears as a reflecting area.
Create a bubble:
Extract some water with the syringe or the eyedropper. With the syringe
let a drop fall on the surface of the water. This creates a bunch of tiny
bubbles. Some of them will dissolve whereas others will drift to the center
of the flask and unite. To see them, you have to light them from behind
and look against a dark background. A broadened laser beam is really nice
and helpful! Watch the microphone signal; Now there should be some ripples
on the sinusoidal signal. The ripples are still visible if the bubble
is to small to be seen without a laser; So watch the ripples, not the bubble.
A high-pass filter improves their visibility. If you see no ripples, increase
the voltage and try again. If the ripples quickly vanish, turn the voltage
lower and try again.
Slowly increase the driving amplitude. At a certain amplitude the bubble
becomes instable and vanishes: No ripples are visible. Turn the voltage
lower a bit and create a new bubble.
This is the big moment:
Darken the surrounding light and look at the center of the flask. You should
see a tiny blue-white dot, like you captured a little star from the night
If no glowing dot is visible: The microphone signal has to give a stable
image of the ripples on the scope. Change the driving voltage and perhaps
the frequency a bit to achieve that. Increase the driving voltage just
below the instability border. You may change the frequency, too, but it
isn't that critical.
Once you succeeded, you should be able to reproduce SBSL without difficulties.
With this basic setup you may start experimenting. Investigate the parameter
range where SBSL is visible. Try other flasks, other liquids (putting glycerine
into the water improves SBSL),...
Persuade many bubbles to glow:
- Fill the flask with water up to the neck.
- Find the acoustic resonance.
- Adjust the inductance.
- Increase the driving voltage until you hear a horrible screeching noise,
which sounds like your flask is going to crack. Don't be surprised if it
does... and don't touch any exposed connection. The sound is produced
by cavitation, i.e. by imploding cavities in the water.
- Pay attention not to overheat the coil(s) or transducers.
- You need a very high driving amplitude (maybe some hundred volts).
If your amplifier doesn't reach it:
To double the voltage, feed the inverted signal to the second (stereo)
amplifier channel and use the signal between the two live outputs.
- Darken your room completely. Adapt your eyes, which may take
15 minutes up to half an hour. Look at the flask or at least where
you think your flask is (remember: the room is completely darkened).
You will see a weak glowing in some areas of the water and sometimes flashes
crossing the flask. This is the famous Multi Bubble Sonoluminescence.
- R. A. Hiller, B. P. Barber, Scientific American Feb. 1995, 78
- D. F. Gaitan et al., J. Accoust. Soc. Am. 91, 3166 (1992)
- L. A. Crum, Physics Today Sept. 1994, 22
- Channel Industries, Inc.
839 Ward Drive
Santa Barbara, Calif.
tel. (805) 967-0171; fax (805) 683-3420
- PI Ceramic GmbH
D-07589 Lederhose, Germany
tel. (+4936604) 8820; fax (+4936604) 88225
*** you may take the PIC 151 or 155 type ***
Zurück zur Homepage
/ back to the homepage