Calibration variation using the 

Radioear B-71 and B-72 Bone Vibrators

Audiologists need to be aware that using a +/-3% frequency tolerance for generating and calibrating reference bone conduction thresholds is prone to variability.  To minimize test variability for bone conduction output during audiometer calibration, a test signal that is as close to the center frequency as possible is desired. 

Audiometer manufacturers are now including the Radioear B-71 or B-72 bone vibrators as standard equipment.  The B-72 is larger than the B-71 because it contains an additional dynamic mass, which lowers its resonances compared to the B-71.  The graph below was obtained from a study by Richards et al. 1982.  It indicates the frequency response of both instruments. 

From Richard et al. 1982

From the graph three resonant peaks can be seen for both instruments.  Due to the added dynamic mass present on the B-72, the resonant peaks occur at lower frequencies than the B-71. The resonant frequency of the B-72 vibrator is approximately 250 Hz, which is slightly lower than the B-71.  Because of this lower resonant frequency, the output is great for the B-72 at 250 Hz compared to the B-71.   This characteristic of the B-72 allows it to be operated at higher output levels with less distortion (greater dynamic range) compared with the B-71 at the lower frequencies.       

When calibrating the bone transducer it is important to know the frequency response of the instrument and more importantly the characteristics of the low and high frequency slope of the resonances.  As a general rule, when any transducer is operated on or near a steep resonant slope, the repeatability of measurements at that frequency decreases.  A study by Richard’s et al. tested the repeatability of measurements for both the B-71 and B-72 by holding the input voltage constant at 0.1 V-rms and monitoring the output voltage of the artificial mastoid (Bruel and Kjaer, model 4930) while audiometric center frequencies were varied by +/-3%.  The results for this test are indicated in the table below (from Richards et al. 1982).  Notice that for all test frequencies, the output voltage was less than or equal to 2 dB except for the B-72 at 4 kHz which was 6.9 dB. 

from Richard et al. 1982

This extreme value at 4 kHz for the B-72 is consistent with the fact that the frequency response curve for the B-72 near 4 kHz shows a steep slope. 

The study by Richards et al. (1982) used an output range which was varied over +/-3% at each audiometric frequency.  +/-3% was done within the frequency tolerance limits for a pure-tone audiometer according to the American National Standards Institute 1969.  American National Standards Specifications for Audiometers.  ANSI S3.6-1969 New York.  Since the publication of this study, the tolerances have been made more stringent and changed from +/-3% to +/-1% for type 1 audiometers and +/-2% for type 2 audiometers.  Audiologists need to  be aware of this revision when their equipment is being calibrated.           

In summary

Audiologists must be aware that using a +/-3% frequency tolerance for generating and calibrating reference bone conduction thresholds can produce variability in test retest results.  A bone vibrator with a flatter frequency response is desirable for test retest consistency but flattening out the response curve would sacrifice the output capability of the instrument.  Since this is clinically not desirable, Audiologists must keep in mind that it is very important during audiometric calibration to have the test frequency as close to the center frequency.  By doing so, the bone conduction output variations that could influence calibration and testing can be minimized.    

Note:  This site is best viewed using IE 4.X or higher under 800x600 resolution