Instrumentation in Communication Sciences Module
TRANSDUCERSThe learning outcomes of this module are for the participant to be able:
Definitions
TRANSDUCTION is the process of taking one form of energy and converting (transducing) it into another type of energy. And of course, the transducer is what does this. The microphone is one type of transducer. It takes acoustic energy and transforms it into electrical energy. The earphone or speaker transduces the electrical energy into acoustic energy.
Microphones etc.
Microphone Operation
There are two types of microphones we will discuss today, the condensor and the dynamic microphone. First, we will discuss the CONDENSOR microphone, used for making precision measurements.
If you were to look at a condensor microphone after removing its protective grill, you would see a thin metal plate, held firmly to the casing. This is termed the microphone diaphragm. It is light weight, and will readily move when sound pressure is applied to it.
Behind the microphone is what is called a back plate. The back plate has been electrically charged - generally at a very high voltage, such as 200 volts.
Figure 1. Illustration of the components of a measuring condensor microphone. Diagram based upon Bruel & Kjaer illustration (although theirs was in boring black & white).
As the metal microphone diaphragm comes closer to back plate, it changes the magnetic field of the back plate. (The opposite change occurs when the microphone moves farther from the back plate during rarefaction.) This creates an electrical field that has the characteristics of the sound wave. The signal is immediately preamplified, since it is a very minute signal.
There are also DYNAMIC MICROPHONES that operate on a different principle. Here the diaphragm of the microphone is attached to a metal piece, which is suspended inside a magnetic coil. When the metal piece moves, it changes the magnetic field inside the coil, creating a voltage that represents the movement of the diaphragm.
Microphone Directionality
Let's begin this section with a definition of incidence - which is, to my mind, the same as azimuth. This is the angle at which the sound is coming into the microphone.
Figure 3. Illustration of angle of incidence. Here the sound is coming in at about a 45 degree angle.
A microphone that is OMNIDIRECTIONAL is one that responds to sound in all directions. One that is DIRECTIONAL is designed to respond best to sound from one azimuth, generally near 0 degrees.
As long as the wavelength of the sound is low (longer than the microphone size) this basic omnidirectional microphone shown in Figure 1 would respond to sound from any direction. If, for example, the sound were coming from behind the microphone, the push/pull of the air molecule movement would still create a movement of the microphone diaphragm.
If the sound is coming straight towards the microphone, and the sound is high frequency, there is a possibility of an inaccurate reading. This would occur if the frequency is high enough that the wavelength is shorter than the microphone. In this case, you could have reflections of the sound waves and some canceling out of the incoming waves. If it were coming from the rear, the microphone would block some of the very high pitch sounds.
Figure 4. The concentric circles are to represent the sound waves. Each wave line would illustrate the air molecules that are in compression. On the left, the wavelength (distance between the compressed air molecules) is relatively short. The object (which could be a microphone!) acts as a reflector. On the right, the wavelength is longer than the object. These low frequencies tend to "wrap around" the object rather than being reflected.
In figure 5 below, you see how a measuring microphone will respond to high frequency sound coming in from various angles. This microphone would measure high frequencies most accurately when it is aimed 90 degrees from the sound source, sometimes called a grazing incidence.
Figure 5. From Bruel & Kjaer Instruction and Application Manual for 4144, 4145, 4156 microphones.
This should illustrate that the high frequency sounds are most susceptible to change in frequency characteristics as a function of sound incidence angle. Is it any wonder, then, that when discussing microphone directionality, a high frequency is chosen as the stimulus. Below are two "polar plots" of hearing aid microphones that use a 2000 Hz stimulus.
Figure 6. Polar plots showing the directional characteristics of hearing aid microphones for a 2000 Hz signal. This is borrowed from Pollack's 3rd edition "Amplification for the Hearing Impaired." (Not to be confused with the book "Amplification for the Normally Hearing.")
Now, this is not to say that smart engineers can't design methods of canceling low frequency noise coming from the rear. They can. It takes more than shielding the microphone. Generally, two microphones are used. The signal from the rear-facing microphone is treated as noise, and this is cancelled from the sound received from the front-facing microphone. So, sounds from near the microphone are picked up, sounds origniating farther from the microphone are cancelled.
An aside on close and far range microphones
As voice command and speech recognition
software become more common place, you will see more and more
headset microphones. These are designed to respond to signals
that are close range - that is within several inches from the
speaker's mouth. How do they do this? The pressure difference
on the front versus the rear of the microphone will be considerably
different for sounds originating near the microphone. The microphone
is designed to respond to these sounds, but cancel out sounds
that have the same (approximate) pressure on both sides of the
microphone. Sounds originating from farther away will have about
the same pressure on each face of the microphone.
Where to place a microphone to record the voice
Where should you
place a microphone in order to optimally record a voice? According
to manufacturer's instructions, of course. If none are supplied,
here's the deal. Ideally, the microphone should be close enough
to adequately record the high frequency sounds, which will fade
with distance. But you don't want to pick up breathing noise,
so you probably don't want the microphone too close to the front
of the mouth. If it is a head-worn mic, then put the mic at the
corner of the mouth. Otherwise keep it about 6 inches away from
the mouth, so the turbulence from mouth breathing doesn't interfere.
Also, you don't want to hear the snap, crackle, pop of the plosives.
Six inches away will permit that to go unnoticed. The best advice
I can give is to listen to the recording quality and be sure it
is free of breathing and mouth noises.
The grills and foam covers for microphones aid in reduction of breathing noises. The turbulence should be diffused before it reaches the microphone.
Accelerometers
Here's one of more concern to speech-language pathologists. Want to see how the jaw moves? The tool of choice is an accelerometer. It has a mass on a spring in it. The mass hangs back a bit once in motion - inertia's effect. Thus, you have a difference in position of the mass and the case. Put some magnetic transducers inside, and like a microphone, the difference in distance can become a difference in voltage.
Pressure Transducers
For measuring air pressure (e.g. intraoral) the transducer can be akin to a diaphragm again. The oral pressure would "blow" the transducer closer to the equivalent of the microphone's back plate, permitting an electrical current to change when airflow changes.
Strain Guages
Strain gauges have a thin piece of wire
stretched taught. When force is applied, the wire will stretch.
When the wire stretches, it doesn't transduce electricity as well.
So, with calibration, we can determine what amount of stretch
equals what amount of decrease in electrical transmission. Thus,
the machine will be able to tell you the amount of strain (e.g.
the amount of movement of the articulators).
Hooking external mics into tape recorders
Tape recorders often have two inputs. One marked "line" and one marked "mic". I'm sure you know enough to put a mic into the mic input, but do you know why?
A line input will assume that the signal is coming from another instrument, and has a certain level of amplification and a certain input impedance (more later). The microphone has been preamplified and has a different impedance.
Transducer Impedance
Having mentioned the term, I had best define it! The impedance of a transducer describes how much it resists the flow of electricity.
Different pieces of equipment can have different impedances. When connecting instrumentation together "in series" (one right after the other, like a microphone into a tape recorder and sending the signal out to earphones) the issue of impedance raises its ugly head. If you don't have equal impedance components, the transfer of energy is inefficient, and you will lose energy. Now, that's not such a terrible thing if you have an amplifier to boost the energy. But bad things can and do happen when you go from a high impedance output of one piece of equipment to a low impedance input of the next piece of equipment. You get distortion!
Field versus Pressure Microphones
Before leaving the topic of microphones, audiologists should be aware of the difference between field and pressure microphones.
Pressure microphones are designed to be used inside a closed coupler, such as the earphone coupler used in calibrating an audiometer's earphone, or inside a hearing aid analyzer coupler. Pressure mics expect the sound to come at a 0 degree azimuth.
If you take the pressure microphone out of the coupler and use it in a free-field environment, your results will not be accurate (for the high frequencies) if you aim it directly at the sound source. You must use the old grazing incidence (90 degrees) to accurately measure the sound source.
Field microphones are designed for free field work, and with this type of microphone you can aim the mic directly at the sound source. You wouldn't want to use a field mic in a coupler. So, if you can only buy one mic, the obvious choice is a pressure mic. Just hang it from the ceiling pointing towards the floor to use it to calibrate the soundfield speakers.
A Few Words on Soundfield Calibration
- Only Audiologists Need Read
Please permit me to give you a few pearls of wisdom I've been given through the years.
Soundfields are messy places. We like to think that putting in a warble tone or a narrow-band noise creates a nice uniform soundfield. In reality, if you experiment with moving the microphone around in a sound room, you will find it just isn't so. Even a clean environment isn't uniform. Dillon and Walker have written on this, advocating much different warble characteristics than the traditional audiometer uses. Putting a human in the booth further screws up the soundfield. But audiologists tend to make matters even worse by putting other things in the soundfield. I've seen tympanometers and tables - tables for play audiometry - and large chairs with plenty of reflecting surfaces. I've heard the rule that there should be no reflecting surfaces within 1/2 wavelength if you want the measurement to be accurate, or the sound to be accurately reproduced.
Let's do the math - the easy way. A 1k signal has about 1 ft. wavelength. One half wavelength is 1/2 ft. Double the wavelength as you halve the frequency, right? So, if you only care about sounds down to 500 Hz, you just need a foot (2 feet wavelength times 1/2) - e.g. between the child's head and a reflective surface, or between the speaker and a reflecting surface. If you would like to measure 250 Hz accurately, you need at least 2 foot separation. If you have that table within 2 feet of the speaker (and heck, the floor is often about as close) you will have problems with the low frequency measurement accuracy. Plus, all reflecting surfaces increase the chances of standing waves. Our warble rates/percentages are not truly sufficient to eliminate standing waves, so our soundbooths should be as sound absorbing as possible.
When you calibrate
a soundfield, it should be free of all reflective surfaces. That's
how the soundfield was designated when they standardized the calibration
method. So, remove the chair from the room - remove yourself as
well. Not always easy to do. Try putting the sound level meter
ontop of the speaker and suspend the (pressure) microphone on
a cable from a bent (large size) paperclip stuck into the hole
in the ceiling. Since the microphone costs $750 - $000, test the
holding power of your paperclip!
I had an equipment calibration service once tell me of their "inventive" trick. They put an inverted trashcan on a chair to get the sound level meter to head height. Not a good idea, particularly when the can is metal and can resonate!
Earphones and Speakers
Transduction Again
Earphones and speakers transduce the electrical energy into acoustic energy. Their operation is essentially the reverse of that of a dynamic microphone. A diaphragm is made to move when voltage is applied.
Inside the speaker is a magnet. Suspended in the middle is the rod, which goes to the cone of the speaker. The cone is held in place by an elastic medium that allows the cone to move.
A speaker's characteristics (frequency response,
power) will be determined by the mass (size) of the cone, and
the size and characteristics of the coil. (A small light tweeter
may be needed in addition to a large heavy coned bass woofer to
cover the full range…)
Types of Earphones
An earphone is essentially a miniaturized speaker. An insert earphone's speaker is even further miniaturized.
Usually earphones and speakers have low impedance - about 4 or 8 ohms. They are matched to the output of the amplifier.
Audiologists may notice that there are different types of earphones, e.g. TDH-39, TDH-49. Each has a slightly different construction, and thus responds a bit differently. The size between the cone and the end of the cushion (MX41/AR audiological standard cushion) can vary a bit, so this means that there are generally slightly different calibration standards depending on the type of earphone.
Some earphones, particularly for evoked potentials testing, use higher impedance earphones, -- 300 ohms. If you put this earphone into a regular audiometer's output, you will have less power. If you go the other way around, a put a low-impedance transducer where a high-impedance transducer should be, you'll have distortion and you'll probably burn out the transducer if you put a lot of sound into it.
Bone oscillators also come in different impedances. There again, swapping transducers may not be a good idea!
When you switch from one earphone/ bone vibrator to another, even of the same type, there are (generally minor) differences in the response of the earphone. Because of this, you can't switch transducers without recalibrating.
For more information visit the Rane Corporation's Web Notes. Note this link will take you away from these course notes. Use the 'BACK' button to return to this page.
Copyright (c) 1999-2009 Nova Southeastern University
Last Updated: