Microphone sensitivity and differential voltage.

[PureBasic]

Hello folks,

Long time I haven't posted there.
I'm stuck in my search for a norm concerning balanced line gains. I remember having seen schematics where the driver is simply reversing one line (leading to a gain of 6dB), some receivers are dividing by two, some receiver are simply making the difference (leaving any gain as is)... It is fine when you go through a device, because as long as the input and output have the same topology, they are transparent gainwise.

The case of microphones is different. The sensitivity is given in dBV (from the factory calibration) but I cannot figure out if this tension is meant to be the difference between the lines or the single ended signal.

 

[user 37518]

Most microphones have an output transformer, that is a floating output (not ground referenced, meaning that NONE of the two pins (2 or 3 in the XLR) needs to be connected to ground), so the voltage is developed between pins 2/3 of the XLR connector, if you stick a multimeter between those pins you'll be measuring the mic output.

So if a mic has a sensitivity of say 10mV/Pa it means that if you capture a sound that is 94dB SPL, the mic output between pins 2/3 will be 10mV, it doesn't matter if you ground pin 2 or 3 (not both of course) of the XLR the voltage between pin 2 and pin 3 is the same, 10mV in this example. Remember voltage is not absolute, it is the DIFFERENCE between two points, ground is just a way to say that it will be the 0V reference for every other voltage in the circuit, and you can make ground any point you want in the circuit.

In the case of a line output there are basically 2 options: transformer balanced or electronically balanced. A transformer balanced is the same as the case of the mic I described, the electronically balanced option has 2 separate sources, one on pin 2 and another on pin 3, one of the sources has its polarity flipped. The important point is that both sources are NOT floating, meaning one end of each source is connected to ground, so if you ground pin 2 or 3 you are basically grounding each side of the voltage source, you are effectively shorting both sides of the voltage source together, thus cancelling that voltage (and creating a whole lot of problems), and you loose all of the signal through that conductor, which is why if you connect electronically balanced to unbalanced you loose 6dB. Now, there is an ingenious circuit known as cross coupled output, in which if you ground one of the lines, the other line raises its output by 6dB to emulate a transformer output so you don't loose level when connecting balanced to unbalanced.

I hope this makes sense, I can post some pics to better illustrate these concepts if you like.

 

[PureBasic]

Oh yeah it totally makes sense. Thanks for the refresh about transformer outputs.
Most mics have discrete output tho (Schoeps style PNP pair or an AB drive). Xformers output are quite becoming the rarity.

I think I'll get myself a calibrated 94dB source to be sure about my measurements. Better be safe.

 

[user 37518]

Oh yeah it totally makes sense. Thanks for the refresh about transformer outputs.
Most mics have discrete output tho (Schoeps style PNP pair or an AB drive). Xformers output are quite becoming the rarity.

I think I'll get myself a calibrated 94dB source to be sure about my measurements. Better be safe.

Not a rarity in dynamics or ribbons...

P.S. Regarding the 94dB source, I bought one from China, it wasn't cheap but a lot cheaper than a B&K or similar. Only problem with those is that they are calibrated at sea level and if you live in a high altitude city (such as myself) the output level will change a bit, for most work this wont make a real difference but if you want precision you'll need to invest in it, I wanted to compare my Chinese source with an expensive B&K from the lab in which I am doing my PhD, but with COVID this hasn't been possible  :-\

Also, remember that those sources are usually for 1/2'' and 1'' mics, so they wont work with something bigger than that.

 

[PureBasic]

I would be interested in seeing the calibrator quality check results also. Aren't those supposed to have a feedback loop and 'servo' themselves to 94dB?

 

[abbey road d enfer]

The case of microphones is different. The sensitivity is given in dBV (from the factory calibration) but I cannot figure out if this tension is meant to be the difference between the lines or the single ended signal.

Tension is the french word for voltage.
You must remember that it is also called "différence de potentiel", which implies that it refers to something that exists between two points. Now, do these points need to be referenced to anything. No. Voltages exist in aeroplanes, so not connected to earth in anyway.
The voltage corresponding to sensitivity is the voltage between two terminals, usually called cold and hot, or even pin3-pin2.

 

[PureBasic]

Hey thanks abbey.
Yeah it is basic stuff. I shouldn't but somehow I needed a confirmation from other people.

I also reverse engineered the measurement mic in question and the output is impedance balanced. So yup. Differential it is.