Implementing reciprocity calibrations inhouse?

Question

Inspired by this question I began to wonder how implementable an inhouse reciprocity calibration would be. In reciprocal calibration, 2 or 3 mics are used as sound sources and as recovers to quantify their sensitivity at various frequencies.

The (perhaps very naive) motivation is that calibration mics are expensive and therefore can't be bought en masse for multiple field sites/personnel.

I found this NIST paper describing the reciprocal calibration workflow, but am not familiar with the detailed electronics and equipment described.

My question is more concretely, do you think it's easier/cheaper to implement a reciprocal calibration workflow for mics than to invest in expensive calibration mics?

Answer 1

For reciprocity calibration one needs 3 devices

• sensor to be calibrated
• sound source
• device that can receive sound AND also emit sound.

For the last reciprocal device the ratio between receive and transmit sensitivity (reciprocity parameter) must be known. This ratio is different for the type of waves used (spherical, cylindrical, and planar waves)

The procedure is rather simple:

• emit sound from sound source, note driving current and measure received voltages on both the sensor in question and the reciprocal device.
• emit sound from reciprocal device with same driving current and measure voltage on sensor in question.

The sensitivity of the sensor in question may then be estimated.

Now specific to OP question: I like the the idea to calibrate a sensor without need for expensive calibration equipment, but the uncertainty I see in knowing the correct reciprocity parameter, let me suggest to spend the money on calibrated sensors for a direct one-on-one comparison