Table of Contents
Advertisement
Amplifiers
Recommended Microphones
Microphones with a sensitivity of 50 mV/Pa (B&K 4189, 4190)
are recommended. The GX-1 is based on a microphone with a
sensitivity of 50 mV/Pa, and corrects the actual sensitivity of
the input system including amp and cable. The limiting of the
sensitivity to 50 mV/Pa was due to such considerations as the
S/N (signal-to-noise) ratio in recorded data and calibration pre-
cision. When a low-sensitivity microphone is used, the S/N
ratio becomes a disadvantage, and piston phone calibration
might fail.
How to Calibrate
Calibrate each channel whenever the GX-1 is powered on.
At piston phone calibration, a special recording method (dif-
ferent from the usual recording method) is adopted. The cal-
ibration is not affected by the current settings for Range,
Sample Frequency, or Filter in the Params window. So there
is no effect on the calibration even if the range is set to a
value below the sound pressure issued by the piston phone,
and even if the sampling frequency is set to a value below
the discriminable frequency of the piston phone sound, and
even if the filter is set.
There are various types of piston phones, but piston phone
calibration of the GX-1 is limited to the following: 250 Hz to
1 kHz, and 94, 114, and 124 dBSPL.
The GX-1 piston phone calibration is performed at a sam-
pling frequency of 20 kHz, so calibration is not possible with
a frequency for which a peak cannot be detected at 20 kHz
sampling.
Calibrate with the post trigger turned off.
To calibrate, attach the piston phone to the microphone, and
turn on the piston phone switch. In Piston Phone, select the
piston phone output, and then click Exec. When calibration
finishes, the correction value is displayed in Coefficient and
the calibrated sound-pressure waveform is displayed. If cali-
bration fails, an error message is displayed indicating that a
peak-detection failure occurred.
The default value 1.0 for Coefficient is the value when the
microphone sensitivity was set to 50 mV/Pa. After calibra-
tion, the display shows the value that is inversely proportion-
4-20
al to the actual sensitivity of the microphone that includes
the amp and cable.
When using a high-sensitivity microphone higher than 50
mV/Pa (including the amp and the drop across the cable), the
correction value becomes less than 1.0; and when using a
low-sensitivity microphone lower than 50 mV/Pa (including
the amp and the drop across the cable), the correction value
becomes more than 1.0. Therefore, you can determine
whether piston phone calibration was performed correctly.
If you note down the correction value, as an alternative to
performing piston phone calibration, you can later type this
value into Coefficient and click OK. However, you cannot
expect precision with this method because the microphone
sensitivity changes depending on such factors as tempera-
ture, humidity, and barometric pressure.
If no piston phone is available to you, as an alternative to
performing calibration, you can obtain (by dividing 50
mV/Pa by the actual sensitivity) the correction value from
the test sheet data attached to the microphone, and type this
value into Coefficient. However, you cannot expect preci-
sion with this method because this value will differ from the
calibration of the entire input system including the amp
amplification factor and the drop across the cable. In this
case, consider the value to be just an estimate.
The correction value is a temporary numeric value used at
calibration, and has no relation to the SLOPE value of the
header file.
Input Range
You can select the input range from 80, 90, 100, 110, 120, and
130 dB. A value of ±100% of the instantaneous value (peak-
to-peak) for the selected range is set as ±100% of the rating (by
using the AD conversion value, ±25000). When the measured
sound pressure is high, select a large range. When the mea-
sured sound pressure is low, select a small range. When a
small range is selected, the S/N ratio becomes poor.
For the AD conversion value, you can obtain a range of
±131% (by using the AD conversion value, –32768 to
+32767); however, the input margin is about ±120% consid-
ering the amp amplification factor, the sensitivity correction,
and the DSP calculation.
Advertisement
Table of Contents
