# How to Measure the Sound dB Level of Ultrasonic

Discussion in 'Electronic Basics' started by William, Jan 6, 2004.

1. ### WilliamGuest

Anybody can tell me how to do the sound level power of ultrasonic sound in db?

Plese help!!!!

2. ### Bob MastaGuest

If you mean SPL, that's normally referenced to 20 micro-pascals, which
is roughly the human threshold of hearing. (I don't know how relevant

that would be to ultrasound, but it is a standard at least.)

So first you have to measure the sound pressure, which requires a
calibrated microphone. You don't state what ultrasonic frequencies
are involved; you can get condenser mics up to about 100 kHz from
B&K or ACO, but expect to fork over several thousand bucks for mic,
preamp, and power supply. They will supply a calibration curve with
the mic. Measure the RMS voltage from the mic, apply the supplied
calibration factor to conver to Pascals, divide by 20x10^-6, take the
log10 and multiply by 20 to get dB SPL.

Note that even with the proper mic, measuring sound at high
frequencies is definitely non-trivial. Every little position change
in your test setup can affect the results, especially if this is
in a reverberant or closed chamber with standing waves.

If you are talking about much higher frequencies than 100 kHz,
you may want to look into thermal measurements (called
bolometry) of sound impinging on materials, etc.

Bob Masta

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com

3. ### aaaGuest

What about if we want to make ordinary sound level measurement,
i.e. in the normal hearing range, do we do it as follows:

We can use an electret microphone connected to an op amp
amplifier, and then we can read the output of the op amp.
I assume the voltage is directly and linearly proportional
to the sound level, so all we need is to measure two
known sounds, we can use an already calibrated sound meter
for this. Then we draw a line between the points and we can
know which voltage corresponed to which sound level.

The only problem here is that we have skipped the frequency
weighting, so we must (if at all possible) choose a frequency
for the calibration which is not affected much by the weighting
curve in the real sound meter (I don't know if such a frequecy
exists) and, of course, then, our meter will be good only for
that frequency.
Is this correct?

JOHNNYbgoode814 likes this.
4. ### Bob MastaGuest

This basic approach will work. (It's the general method for
calibrating mics from a known reference mic.) Most sound
level meters have a 'Flat' or unweighted range, which you
should use. Electret mics tend to have a fairly flat frequency
response if they are the "omnidirectional" type; "unidirectional"
types sacrifice flatness for directionality.

At high frequencies the sound field is easily disturbed, so it
may take a bit of fiddling to get a test setup that allows both
mics to read the sme sound field. You might try pointing the
mics at each other a short distance apart, 90 degrees from
the incident sound. Or you might need a fixture that allows
you to make readings with one mic and then replace it
exactly with the other and repeat.

At ultrasonic frequencies, it is not only hard to find a flat
mic response, it is even harder to find a flat source,
especially since there is such an interaction with the
surroundings. So you must use care when deciding
what frequencies to use for calibration. Make sure they
don't fall in the response notches of either mic or the
source, and hopefully not on a steep slope. (The danger
is that these things change with temperature and small
position changes, so repeat readings give wildly different
values if you are on a steep slope or notch.)

Bob Masta

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com

5. ### JoeGuest

If I may, I have been working on a sound level meter in recent days. I
wanted to build one to check the sound levels of some of the ultrasonic pest
repellers on the market. I received one from someone and thought I would
test it and see what levels it outputs. I don't need anything super
accurate, just something that can 'hear' and measure what I cannot with my
own ears. FYI, I am a hobbyist, not a design engineer.

I am using a 3 and 3/4 inch tweeter that is specd at 3800 to 30Khz as my
input transducer. I have noticed that it is VERY sensitive at 5Khz (about
10dB up), but pretty flat response up to 20Khz.. I also came across an old
article by forrest mims where he discussed this topic. He was using a radio
shack sound level meter in his discussion. He gave some data on different
sound sources at distances, and suggested that an electric razor, or a home
vaccuum cleaner is an excellent source of a broadband of sound. I am using
the vaccuum cleaner at 2 meters to calibrate my meter. A vaccuum cleaner
seems to be the only thing my dog will run away from with his hair standing
on end, so I am thinking there must be some ultrasonic frequencies present.
According to his article, It would be approximately 75dB at 2 meters
distant. I was planning on tweaking my amplifier to give me about a 3/4
reading on my output analog voltmeter so I would still have some 'room' for
higher sound levels. I don't need a whole lot of precision. I am just
looking for relative loudness of these pest repellers. I don't know if this
helps, but just thought I would let you know. My project is more of an
experiment right now, but I may get some useful information from it
eventually. I am measuring the responses of the tweeter and my circuit on a
tek 465B oscope.

Conceptually, what I have done is used the tweeter as a transducer, coupled
to a TL082 opamp with a gain of 50. That signal is coupled to the other half
of the TL082 with a gain yet to be determined (around 20 seems about
right--but still TBD). Then my plan is to rectify the output signal and
display it on an analog voltmeter. I know that my capability is going to be
limited to 30Khz because of the transducer, but I couldn't find an
inexpensive broadband ultrasonic transducer to work with.

Joe