Low Polarizing Voltage = Decreased Sensitivity?

[hodad]

I feel kind of stupid for not knowing the answer to this, but would a too-low polarizing voltage on a condenser capsule cause low output level?  The mic in question is an old Sony C38, which is very similar to, if not the same as the C37 FET (manual here:http://www.coutant.org/sonyc37fet/c37fet.pdf).  The output is extremely low, but not noticeably distorted.

It runs on battery only, no phantom powering possible.  Polarizing voltage is via a DC-DC converter.  According to the manual the polarizing voltage can't be checked with a normal voltmeter (though possibly with modern digital multimeters it wouldn't be a problem?) 

Am I thinking right that low polarizing voltage could be the source of this problem (not that I'm excluding other possibilities at this point)?  Or am I just wrong?

 

[gyraf]

yes, lower pol. voltage > lower sensitivity

 

[hodad]

Thanks. After I've eliminated a few more basic possibilities, I'll look at the DC-DC converter.

 

[PRR]

Think of the extreme. If the diaphragm moved 100% of the way to the back-plate, the output would be 60V peak. Or if polarizing voltage were 10V, the same motion would give 10V peak. With zero polarization, zero output.

Obviously normal motions are under 1% of backplane distance. Also the induced voltage is not that neatly linear. But yeah, normal output is some pretty direct function of polarizing voltage.

> polarizing voltage can't be checked with a normal voltmeter (though possibly with modern digital multimeters it wouldn't be a problem?)

There's no real difference between the 11Meg input of VTVMs found in service 1950-1980, and the 10Meg input of most DMMs used recently. The polarizing supply only has to supply leakage, which must be many hundreds of Megs, and they may not build any excess capacity into it. A meter load 1/10th of normal load will probably sag it, even stall the inverter.

However DMMs do resolve milliVolts better than bench VTVMs. You could rig a 1Gig:1Meg voltage divider. 60V in gives 60mV out, minus 10% for 10Meg meter loading, around 54mV. This assumes the inverter does not sag much with 1Gig load.