Is there a standard way to express a noise level

I am confused. The measured noise level of equipment or components seems to lack a standard. EG: an audio power amp is said to have a noise level of -85dB, compared to what?. I have seen such things as "below rated output" fair enough ,you inject a tone at a level which resuts in say 50W into an 8R load then remove the signal and measure the noise. The two will have a voltage ratio which can be expressed in dB's. So far so good but the the other day I am searching for mic capsules and I find an electret that has an SPL of 130dB and a noise level of -66dB. Is that purely it's self noise including the FET or was it measured by applying an audio signal at a specific frequency etc at a level which produced a certain output and then removing the signal and measuring the noise.
IS there a standard and if there is has anyone told the manufacturer.
Steve

For microphones, the standard noise measure is the "self-noise" measured in absolute dBA. This is is a measure of what equivalent ambient noise level the microphone's self-noise equals to.

Note that this measure is independent on the capsule sensitivity or built-in amplifier - making it easily cross-comparable.

Most electrect capsules are not specified with regards to this standardized self-noise, simply because it would be very clear how noisy this technology is compared to other types.

Jakob E.

Jakob,

Do you mean that the electret principal is inherently noisy or the capsule including the FET.
Steve

Just to tag on to what Jakob has said here...


Look here on the Rane website...scroll down to the short explanation of Self-Noise. So this test would obviously be done in an anechoic chamber, where ambient noise is pretty close to nothing.

Looking at some Shure spec sheets, they spec "IEC-651" so there is some particular way that this test is supposed to be performed, I'm certain.

HTH!

> Is there a standard way to express a noise level

No. Then you could do honest comparisons.

The several Standards Agencies are part of the problem. Hi-Fi electronics is often referenced to arbitrary levels, 1 Watt for power amps.

> I find an electret that

The Panasonic capsules are rated re: 94dB SPL, a loud speaking voice. So 66dB means self-noise is less than 28dB SPL. This is quieter than many residential rooms, but higher than a good studio. My observations say that the P-Sonic capsules are mostly much better noise than their spec, and that there are a few bad apples in the barrel. Instead of using a proper grid/gate resistor they depend on leakage, which is very variable, so maybe 95% will be pretty good and 3% will be hissy (and 1%-2% get tossed at the factory when they pin the hiss-meter).

That 130dB spec must be maximum for low distortion. That will cover all sane music. It will not cover close-drums, and is mighty marginal on close guitar-amp (though you may like the sound as the capsule overloads).

An other dual ansver (because I love duality)

Is there a standard way to express a noise level

No. Then you could do honest comparisons.

Click to expand...

Yes, TEMPERATURE because of thermodynamic laws.
In every systhem system at thermal equilibrium
(mostly at the some temperature)
there is not direction of energy flow.
(In opposite case, energy flow can be transformed and we have
perpetuum mobile)
In normal studio we have laboratory
temperature. If we know all equivalent scheme of microphone
and amplifier, only recompute for noise with
studio temperature.
Some problem is, thet no all noises are thermal,
but it is not problem in acoustic and transducers, it is
problem of electric circuits and here noises can be determined
separately.

Sometime, to compute noise is simplier than to direct measure,
and if your model is realistic, it is precise.
And for partial measurements you need only set of resistors and
FFT analyzer.
Resistors with no DC current are all enough precise, because of perpetuum mobile.
Note: Use Motchenbacher rather than SPICE.
xvlk