Sorry if this bores any of you, I know it's getting rather technical.
Here's how it's done. You need a 150 ohm resistor, a sound editing program
with good dB scales or analysis abilities like Audition, and you must know
your mic preamp gain.
Place the resistor across pins 2 and 3 of your preamp/recorder (turn off
phantom power). Turn the mic trim (level controls) all the way up (so they
are not attenuating the signal at all) and record the faint noise. If you
don't have the resistor then use a wire or leave it open, all the
measurements will be close, but only 150 ohms is the standard.
Pull the recording into the audio editor. If you have a way to limit the
bandwidth to 20-20k exactly, do so. If you can A-weight the audio you might
do that too. I built a 23-20k FFT filter to do this, flat-topped in the
passband. Play the noise, and watch the meters to estimate the level of the
noise.
To determine the EIN, you must add the noise just measured to the gain of
all stages prior to recording. The result will be in dB, but I don't know if
it's possible to put the dB's on a specific reference, like dBu. To do that
you'd need a very good rms voltmeter to measure the output of the preamp
stage, then you could put the noise measurement on an absolute scale.
On my new 722 (woo hoo!) using a wire (not a resistor) I get a peak reading
of -55 dB FS. In the analysis menu I get an average of -66.7 RMS. If I trust
the 70 dB gain figure quoted in the 722 literature, then the EIN is -125 dB
peak, and -136.7 dB average. This is compared to the thermal noise of a wire
at room temperature of -145 dBu or so. Close to the quoted figures of -131
dBu for the unweighted 50 ohm EIN.
For my older 670, the noise measurement (using the wire) is -54 dB
peak, -59.3 dB average rms. Using the 48 dB gain value calculated form the
1.2 mV MIC to 300 mV LINE difference in the specs, the EIN is -102 dB
peak, -107.3 dB average. This is quite a bit above my calculated value
of -121 dBu.
If you don't know the preamp gain then you'd need to measure it by inserting
an audio tone of known rms voltage into the mic input, record at full volume
(zero attenuation) and observe where it ends up during playback. You could
then calculate the gain. You could then calculate the EIN in reference dB
units, like dBu or dBV.
I need to go get a 150 ohm resistor now.
Bruce Wilson KF7K
http://science.uvsc.edu/wilson
________________________________________________________________________
________________________________________________________________________
|