How to measure EIN (Equivalent Input Noise) of an interface's Mic preamp

[nanook]

Well, if you want to get accurate, reproducible result within 0.1dB or so, you need to correct it,

I agree, it's almost no effort to correct it.

When I worked for a company developing low noise measurement amplifiers, we never corrected for the exact temperature because the unit to unit (OpAmp to OpAmp) variations swamp this anyways.

 

[nanook]

Isn't the amount of S/N degradation a device introduces the thing we really want to know? Noise figure is exactly that amount.

Although it's a term originating in the RF amplifier area, this is exactly what is describing an amplifier's noise performance best for a first glance view.
(Albeit we had some discussions on how to calculate it although it seems to be pretty straightforward.)
And isn't the difference between the calculated EIN and the EIN calculated for a noiseless preamp (source resistor noise only) already close to the noise figure?

Audio doesn't live in the standard 50 Ohms world, so we would need to have a number at different source impedances, e.g. 0 Ohms, the usually specified impedance 150 Ohms, 600 Ohms and 10k. Otherwise we wouldn't take care of the current noise.

The noise densities en and in have a 1/f character with the noise corner usually being somewhere between 10 Hz and 1 kHz (bipolar usually lower than FET based inputs). So this may be important as well.

If we would extract the voltage- and current- noise densities en and in at a few frequencies everyone can calculate the final EIN with the source impedance he's expecting.
This would be my favourite approach.

Here's a nice summary on EIN for those not yet really familiar with the term:

What Is Equivalent Input Noise (EIN) In Audio Interfaces?

Uncover the concept of Equivalent Input Noise (EIN) in audio interfaces. Learn its measurement, significance, and how it impacts pre-amp noise levels.

audiointerfacing.com

 

[KSTR]

And isn't the difference between the calculated EIN and the EIN calculated for a noiseless preamp (source resistor noise only) already close to the noise figure?

Yes. But that ideal EIN must be calculated/given.
And current noise is an issue, though mabye not so much for mic pres which are driven from lower impedances.

In the end we're back to @SIY proposal to measure current and voltage noise input referred noise densities which has the drawback of not being so easy to do for most people that @Rja4000 wanted to address to take measurements.

----:----

Here is a doc about the relationship between "Noise Figure and Equivalent Input Noise" (exact document title):

 

[Blumlein 88]

I like doing these sorts of things, but when looking at Julian Krause's data I ran across one of his old videos I've seen, but forget about. It shows you why with condenser microphones EIN does not matter. The self noise of even quiet microphones swamps the EIN of all, but the worst mic preamps. If EIN is at least -120 dbu, it is a pointless exercise. It can matter with ribbons and dynamics.

 

[nanook]

It shows you why with condenser microphones EIN does not matter. The self noise of even quiet microphones swamps the EIN of all, but the worst mic preamps.

That's the voltage noise of the JFET in the capsule / body of the microphone that does the impedance conversation (and sometimes gain).

 

[Blumlein 88]

Here are the claimed specs of the Earthworks ZDT microphones. Includes the noise density spec at 20 db, 40 db and 60 db gain.
I notice the input impedance is 10 kohm with phantom power on and 100 kohm with it off.

https://www.earthworksaudio.com/wp-content/uploads/2013/08/ZDT-Series-Data-Sheet.8.13.pdf

 

[Rja4000]

Here are the claimed specs of the Earthworks ZDT microphones. Includes the noise density spec at 20 db, 40 db and 60 db gain.
I notice the input impedance is 10 kohm with phantom power on and 100 kohm with it off.

https://www.earthworksaudio.com/wp-content/uploads/2013/08/ZDT-Series-Data-Sheet.8.13.pdf

View attachment 388585

I need to find one of those for measurements

 

[Blumlein 88]

I need to find one of those for measurements

It's possible a pro audio place might have one to rent depending upon where you live.

 

[Rja4000]

It's possible a pro audio place might have one to rent depending upon where you live.

I'm in Belgium
I'll do a search.

 

[nagster]

I need to find one of those for measurements

I have a 1022. I can't send, but would a simple measurement using APx555 help? If so, please tell me what measurements you need and what conditions you need. I think I can measure it within 2-3 days. But I can't do a lot.
If need the method in post #1, it will be a while later. I've been busy lately...

 

[nanook]

Here are the claimed specs of the Earthworks ZDT microphones. Includes the noise density spec at 20 db, 40 db and 60 db gain.
I notice the input impedance is 10 kohm with phantom power on and 100 kohm with it off.

https://www.earthworksaudio.com/wp-content/uploads/2013/08/ZDT-Series-Data-Sheet.8.13.pdf

View attachment 388585

The voltage-noise-density and the calculated EIN values must have been measured with the input shorted. They should have noted this condition.
(a 22 Ohm resistor has a voltage noise density of 0.6nV/sqrt(Hz) )

There is a nice calculator online: https://www.beis.de/Elektronik/Nomograms/R-Noise/ResistorNoise.html
(It does not apply A-weighting, so the numbers will differ)

 

[Blumlein 88]

The voltage-noise-density and the calculated EIN values must have been measured with the input shorted. They should have noted this condition.
(a 22 Ohm resistor has a voltage noise density of 0.6nV/sqrt(Hz) )

There is a nice calculator online: https://www.beis.de/Elektronik/Nomograms/R-Noise/ResistorNoise.html
(It does not apply A-weighting, so the numbers will differ)

Yes, I noticed that. I assumed they were only showing the noise from the preamp itself.

 

[nanook]

Yes, I noticed that. I assumed they were only showing the noise from the preamp itself.

That's why I like the noise spectral densities (en, in) rather than EIN.

EIN offers a temptingly simple number but still you have to look carefully at the conditions.
- it needs the source impedance to be specified
- it may be given in dBu or dBV
- it may be A-weighted or not
- it would not really reveal a 1/f corner as high as e.g. 1 kHz (because this represents only 1 kHz bandwidth out of ca. 20 kHz)

The spectral densities en, in on the other hand need the frequency to be specified; usually they're specified at 1 kHz. Above that they should definitely be flat and hardly anyone will voluntarily specify them at lower frequencies.

 

[Blumlein 88]

That's why I like the noise spectral densities (en, in) rather than EIN.

EIN offers a temptingly simple number but still you have to look carefully at the conditions.
- it needs the source impedance to be specified
- it may be given in dBu or dBV
- it may be A-weighted or not
- it would not really reveal a 1/f corner as high as e.g. 1 kHz (because this represents only 1 kHz bandwidth out of ca. 20 kHz)

The spectral densities en, in on the other hand need the frequency to be specified; usually they're specified at 1 kHz. Above that they should definitely be flat and hardly anyone will voluntarily specify them at lower frequencies.

I agree, but at least for as long as I can remember you have a heck of a time getting people to show specs with the pertinent parameters. Look at all the microphones that never list more than Response: 20 hz to 20 khz. Yeah, sure at what level and within what range etc, etc. Not worth even listing if that is all they include.

 

[Rja4000]

If we agree to standardize on EIN unweighted with shorted inputs, 20Hz-20kHz, we have your Noise density and a readable value.
We may add A weighting as well and then to add a computed EIN 150 ohm 20°C in dBu for comparison.

 

[nanook]

If we agree to standardize on EIN unweighted with shorted inputs, 20Hz-20kHz, we have your Noise density and a readable value.
We may add A weighting as well and then to add a computed EIN 150 ohm 20°C for comparison,

I agree, it would be nice to have a number all users can "read" and compare.
EIN will not be easily comparable to manufacturer specs though. Among the ASR measurements comparability is given of course.

- manufacturers will try to state low numbers, so A-weighting will be the usual thing and for Audio it makes perfectly sense
- When you spec in dBV you get a lower number, although dBu rather is "audio"
- manufacturers
- units with a unusual high current noise would look better than they are (although with 150 Ohms the voltage noise will be dominant in most cases)
- some manufacturers will use a lower source impedance than the usual 150 Ohms (and not even note this)

How about measuring and stating noise densities and having an online calculator that allows to calculate the EIN for the conditions you want to compare to (Zsource, w/wo A-weighting, dBV/dBu)? This calculator could even give the number for a noiseless preamp for comparison.
I must however admit, that I wouldn't be able to write the code for such a calculator.

 

[KSTR]

Isn't that all in the paper I linked?

 

[Rja4000]

I have a 1022. I can't send, but would a simple measurement using APx555 help? If so, please tell me what measurements you need and what conditions you need. I think I can measure it within 2-3 days. But I can't do a lot.

That would be really nice !
Could you perform 3 measurements ?

1. Max gain in dB

2. rms Noise at the output (output level @0dB) at max gain in V or dBu, 20Hz-20kHz, unweighted, input shorted

We can then deduct EIN, which is Noutput - Gain, for an analog preamp

3. THD and/or THD+N sweep vs level at min gain from -30dBu to max level (+30dBu ?) at the output, 61 steps
(10Hz-22kHz, 1kHz sine or similar)

Do you think that's do-able ?

I'm really curious...

 

[Rja4000]

I agree, it would be nice to have a number all users can "read" and compare.
EIN will not be easily comparable to manufacturer specs though. Among the ASR measurements comparability is given of course.

- manufacturers will try to state low numbers, so A-weighting will be the usual thing and for Audio it makes perfectly sense
- When you spec in dBV you get a lower number, although dBu rather is "audio"
- manufacturers
- units with a unusual high current noise would look better than they are (although with 150 Ohms the voltage noise will be dominant in most cases)
- some manufacturers will use a lower source impedance than the usual 150 Ohms (and not even note this)

How about measuring and stating noise densities and having an online calculator that allows to calculate the EIN for the conditions you want to compare to (Zsource, w/wo A-weighting, dBV/dBu)? This calculator could even give the number for a noiseless preamp for comparison.
I must however admit, that I wouldn't be able to write the code for such a calculator.

Well, you don't really need a calculator.
A simple Excel sheet would do.

 

[AnalogSteph]

How about measuring and stating noise densities and having an online calculator that allows to calculate the EIN for the conditions you want to compare to (Zsource, w/wo A-weighting, dBV/dBu)? This calculator could even give the number for a noiseless preamp for comparison.

Weighting and non-flat noise densities can be done but it's a bit of a pain as you basically have to divide your frequency range of interest into a bunch of chunks, each with roughly constant noise density and weighting factor. (People did this for phonopre noise calculation in the '70s.) You not only have to pick the ranges but also enter the noise density as well, which quickly becomes tedious when doing it by hand. If you are happy with a flat noise density approximation, my RMS summing and unsumming calculators should do the job.

If you want A-weighting, I'd suggest using something like REW to determine the difference to flat 20-20k on your measurement spectrum (no need to reinvent the wheel if someone else has already done the work, and it keeps you from having to type a bazillion values). I have also used an A-weighting EQ preset for Audacity's filter curve effect before. (Note that A-weighting is only considered valid up to 20 kHz, and any substantial content above this may need to be disposed of using a sharp lowpass filter.) The wave file analysis of RMAA should also do the job, I think... Plus I reckon there's various metering plugins for your favorite DAW that would work.