• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Microphone preamp measurements: How to get the signal level right

OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
That would mean 400 Ohm input impedance or termination. These are all balanced inputs, so it would be correct to use one 200 Ohm resistor between pins 2 and 3.
@MC_RME
Hi Matthias
It seems most common is either 150 ohm either 0.
For the Quadmic II, you use 150 ohms in your specs.
Is that A weighted or not ?
 
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
What microphone source impedance do you optimize for?
Well, the examples I've put in my introduction are not random.

I have 45 microphones.
The Audix OM7 is at one extreme, with very low sensitivity of 0.8mV/Pa and 50 ohm.
At the other extremity, I have a Neumann TLM103.
Admittedly not the best Neumann around, but its sensitivity is 23 mV/Pa at 50 ohm and the max output level is 13dBu (so, in fact, I guess I could feed the RME ADI-2 directly with it. OK, that's for 138dB SPL. And it requires phantom power.)
 
Last edited:

MC_RME

Addicted to Fun and Learning
Technical Expert
Audio Company
Joined
May 15, 2019
Messages
854
Likes
3,564
What microphone source impedance do you optimize for? A powered mic. that will tend to have a higher output but lower output impedance or a dynamic mic. with an inherently higher output impedance but somewhat lower output and, therefore, desperately in need of the quietest possible gain?

That's not the point. International standards use either 200 Ohm or 150 Ohm to measure EIN. Your proposal to use 2 x 200 Ohm resulted in 400 Ohm effective input termination, so gives wrong results.
 

Pluto

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2018
Messages
990
Likes
1,631
Location
Harrow, UK
International standards use either 200 Ohm or 150 Ohm to measure EIN
Yes I get that but my question remains. What is optimum mic. impedance to design for to obtain the lowest possible noise on low output dynamic mics. that need rather more help in this respect than condensers?
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Dynamic mic noise is pretty much just that of the output impedance. Condenser self noise is above noise from just impedance. Shure SM58 mics have 150 ohm output impedance. As does the SM7b.

Most preamps are adding less than 3 db to that.
 
Last edited:
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
Yes I get that but my question remains. What is optimum mic. impedance to design for to obtain the lowest possible noise on low output dynamic mics. that need rather more help in this respect than condensers?
A few mic I own (dynamic only):
Audix OM7 0.8mV/Pa, 50 ohm
Shure SM58 1.85mV/Pa, 150 ohm
Sennheiser E906 2.2mV/Pa, 350 ohm
Beyer M201 1,2mV/Pa, 200 ohm
Revox M3500 1,3mV/Pa, 600 ohm (similar to the 201, but different coil)

Also, although not a mic:
Radial Engineering JDI Duplex, 150 ohm

And one I dont own:
Shure SM7B 1.12mV/Pa, 150 ohm

...
 
Last edited:

Pluto

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2018
Messages
990
Likes
1,631
Location
Harrow, UK
Dynamic mic noise is pretty much just that of the output impedance
Sorry - I'm obviously not asking my question clearly enough. Given that low output dynamic mics. need the most gain, are we designing amplifiers optimized to produce the lowest possible noise figure with this type of mic, rather than higher profile condenser mics. which, because of their internal amplifiers, are less in need of truly low noise amplification?
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Sorry - I'm obviously not asking my question clearly enough. Given that low output dynamic mics. need the most gain, are we designing amplifiers optimized to produce the lowest possible noise figure with this type of mic, rather than higher profile condenser mics. which, because of their internal amplifiers, are less in need of truly low noise amplification?

EIN specs are normally quoted for 60 db gain or max gain if a device doesn't reach 60 db. Some designs have lower EIN performance as you lower gain. 50 to 60 db is what you'll need with a dynamic. 30-40 db gain is more common with condensors. Good designs will maintain EIN over that amount of gain. So if there is an edge you have a bit better preamp noise in the range dynamics are usually going to be used. If that is what you are wondering about.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
Some designs have lower EIN performance as you lower gain.
Some? I'd say the vast majority of them do. It's just the nature of the beast in typical circuits.
Designing Microphone Preamplifiers - THAT Corporation
Not saying it couldn't be done any differently, but if you need lowest EIN at low gain the feedback networks get lower and lower in impedance and current demands are up accordingly. Basically, you're asking for a higher instantaneous dynamic range (because same output signal but less noise due to less gain), and that automatically comes with higher power demands.

As a example with a simple noninverting opamp circuit:
Gain ~60 dB: Rf = 10k + Rg = 10 ohms = 10.01k. Basically any opamp can drive that to high levels.
Gain ~21 dB: Rf = 100 ohms + Rg = 10 ohms = 110 ohms. That's a way, way tougher load. Time to break out a buffer circuit.

Apparently the M-Audio F6 features a fixed-gain input stage and does the rest with its hybrid ADC setup. Its input gain adjustment is just float32 multiplication.
 
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
Apparently the M-Audio F6 features a fixed-gain input stage and does the rest with its hybrid ADC setup. Its input gain adjustment is just float32 multiplication.
I think the Yamaha AD8HR is doing something similar. That's probably why you have those "steps" in the SINAD vs gain plot.
 
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
Millennia HV-3C as a Phono Preamp ?

Just for fun, I wanted to see if the Millennia HV-3C could be a good (expensive) Phono Preamp

So I've used Virtins Multi Instrument, as I always do, which also has an "RIAA reproduction" compensation curve.


Here is the SINAD at 1V output, 39.7 actual gain, with RIAA compensation

2020-08-13 18_18_51-Greenshot.png


So it's a little less than the best-of-the-class, the Cambridge Audio Duo, which reaches 90dB under same conditions.
(We can get the same 90dB SINAD figure with the Millenia @10mV input if we push the gain to 46dB. We have 2V output then.
At least, that means that my test signal is not the limiting factor here)


Where it becomes funny, is when you try to reach max output with the HV-3C at same gain :

2020-08-13 18_25_11-Greenshot.png


I have no clue if any moving magnet cartridge is able to have 184mV output level, but if there is one, then you get 17.7Vrms at the output of the preamp before saturation (THD <0.0002%!)
That's +27dBu.
More than the RME could swallow.


OK, that was just for fun.

@jan.didden You're a genius !
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Some? I'd say the vast majority of them do. It's just the nature of the beast in typical circuits.
Designing Microphone Preamplifiers - THAT Corporation
Not saying it couldn't be done any differently, but if you need lowest EIN at low gain the feedback networks get lower and lower in impedance and current demands are up accordingly. Basically, you're asking for a higher instantaneous dynamic range (because same output signal but less noise due to less gain), and that automatically comes with higher power demands.

As a example with a simple noninverting opamp circuit:
Gain ~60 dB: Rf = 10k + Rg = 10 ohms = 10.01k. Basically any opamp can drive that to high levels.
Gain ~21 dB: Rf = 100 ohms + Rg = 10 ohms = 110 ohms. That's a way, way tougher load. Time to break out a buffer circuit.

Apparently the M-Audio F6 features a fixed-gain input stage and does the rest with its hybrid ADC setup. Its input gain adjustment is just float32 multiplication.
I should have worded things more clearly.

I've only had hands on a few mic pres to test the EIN. A couple had almost the same EIN from 40-65 db gain dropping off at lower gains. A couple had a good reading at 60 db, but EIN dropped pretty sharply as gain was reduced. Sort of what you are seeing in the graph of Sinad up thread. The Millienia had SINAD nearly the same over a more than 24 db gain range. While the Mackie has a peaked curve which was good only at one gain setting. I'm assuming the SINAD was being compromised by the noise floor from worsening EIN.

I forget who, maybe it was Metric Halo, or someone who claimed to have a mic pre with the same EIN from 20-80 db gain. But that is not common.
 
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
I try to evaluate EIN correctly.
Doing this +/- 0.1dB is a challenge.
+/- 0.2dB seems OK.

It's pretty clear we need it at 150 ohm.
I don't know if that's supposed to be A weighting or not. I guess yes. ( Even if Crane says the opposite)

So I measured with the Millennia.
I'm trying to automate things, and the challenge is the Pre-DUT gain (or attenuation).
I guess I need a good True RMS meter (with PC interface), but even then, 0.1dB accuracy will be a challenge.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
I try to evaluate EIN correctly.
Doing this +/- 0.1dB is a challenge.
+/- 0.2dB seems OK.

It's pretty clear we need it at 150 ohm.
I don't know if that's supposed to be A weighting or not. I guess yes. ( Even if Crane says the opposite)

So I measured with the Millennia.
I'm trying to automate things, and the challenge is the Pre-DUT gain (or attenuation).
I guess I need a good True RMS meter (with PC interface), but even then, 0.1dB accuracy will be a challenge.
For EIN, I wouldn't sweat it past .5 db myself. Even temperature has a small effect. 2 degree celsius is almost .05 db.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
If in doubt, include a measurement at 50 ohms or shorted as well to reduce the effect of errors. I like having shorted EIN as well - it makes telling apart really low noise amps a good bit easier, otherwise they're all kind of bunched up.

Something like 600 ohm may be interesting in case of hot-rod bipolar inputs with substantial current noise, though that's in the region where input impedance really isn't negligible any more. The resulting attenuation alone effectively drives up EIN. Not so much at 3 kOhms but 1.7k or 1.5k is a different story.

Ideally you'd have both A-weighted and defined-bandwidth flat EIN (could be calculated from the same measurements). The former is better at ignoring audibly insignificant low-frequency shenanigans, the latter is better for doing math with (like adding or subtracting the thermal noise of a given resistor). The difference between A-wtd and 20k flat should usually be about 2.2 dB on pure white noise if memory serves.
 
Last edited:
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
Tests progressing slowly...

Millennia HV-3C

Actual gain and channel matching vs gain

Channel matching.png

Left and Right channels have equal actual gain +/- 0.02dB

Actual gain is approx 1dB less than indicated,
but this is very constant accross full gain range.
Max gain is approx 60.4dB, at "61.5" mark.


EIN (A weighted) vs Gain - 150 ohm - 20Hz-20kHz

EIN.png


Under test conditions, measured EIN (A) varies from -128.7dBu at full gain up to -110.3dBu at lowest gain (Lower = Better)
We have EIN < -120dB on a large gain range, from 22.5dB mark up to 61.5dB .
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Tests progressing slowly...

Millennia HV-3C

Actual gain and channel matching vs gain

View attachment 78323
Left and Right channels have equal actual gain +/- 0.02dB

Actual gain is approx 1dB less than indicated,
but this is very constant accross full gain range.
Max gain is approx 60.4dB, at "61.5" mark.


EIN (A weighted) vs Gain - 150 ohm - 20Hz-20kHz

View attachment 78324

Under test conditions, measured EIN (A) varies from -128.7dBu at full gain up to -110.3dBu at lowest gain (Lower = Better)
We have EIN < -120dB on a large gain range, from 22.5dB mark up to 61.5dB .
Yeah I bet that last chart EIN vs Gain will be much less impressive with the other preamps you have on hand. Good work doing this in any case.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
I'd be guessing they're using a digital gain control setup as e.g. provided by some THAT ICs.

A best-case -128.7 dBu(A) EIN is perfectly competent but not actually any better than a number of sub-$200 audio interfaces. (Spec seems to be -133 dBu shorted @ 30 kHz, which should have resulted in -129.4 dBu @20 kHz unweighted or -131.6 dBu(A). Might be time to double-check the generator setup / output noise level. We are talking 0.14 µV @ 20 kHz or the equivalent of a nominal 62 ohms at 295 K for that input. Signal gain vs. noise gain may also be off, though I don't see how this much. Maybe look at spectra to identify possible stray radiation or whatever.)
What makes this unit so outstanding is that it provides a maximum input level of +23 dBu at minimum gain (+8.5 dB nom). So those -110 dBu of EIN correspond to a maximum dynamic range of 133 dB(A). About 8-10 dB of which you may never even use, mind you - +32 dBu is seriously hot. I guess the point is never running out of headroom, or conversely, being able to use higher than average gain.

A look at distortion vs. gain at either maximum output level or something more realistic (like +22 dBu or +10 dBu) may prove interesting. The optimum is likely to be somewhere around the +30 dB point - above that, your loop gain starts to dwindle, and near minimum performance is degraded by high input voltage swing, at least on the more pedestrian gear that I've looked at. It probably is no accident that the datasheet distortion specs are given at 35 dB and +27 dBu out.
 
Last edited:
OP
R

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,681
Likes
4,218
Location
Liège, Belgium
128.7 dBu(A) EIN is perfectly competent but not actually any better than a number of sub-$200 audio interfaces. (Spec seems to be -133 dBu shorted @ 30 kHz, which should have resulted in -129.4 dBu @20 kHz unweighted or -131.6 dBu(A). Might be time to double-check the generator setup / output noise level.
That was the whole goal of this thread: how to generate proper test signal, which basically means, with low enough noise at very low level.
I'm pretty confident my signal is "good enough" down to 1mV, which is low enough to measure up to 80dB gain.
But for EIN, there is, of course, no generator.

For noise measurement:
The setup is
An XLR with R150 1% metal resistor is soldered berween pins 2 and 3 in an XLR connector.
This connector is plugged in the HV-3C input
HV-3C output is plugged in an Y type XLR cable, which feed bot inputs of the RME ADI-2 Pro fs R.

The RME is set for +4dBu input range and M/S mode. The M channel (Sum/average) signal is then used in Virtins Multi Instrument 3.9.

I use an FFT with rectangle window, 65k size, 48kHz frequency and I'm measuring RMS level on 20-20kHz band, with A weighting. I read it after average on 15 frames.

If I plug the 150 ohm resistor XLR to the Y cable, and then directly in the RME, I read -124.7dBFS, which is way lower than any reading through the Millennia, of course.

I spent time trying to measure the matching of Input reading to actual dBu for the RME. I miss a good enough voltmeter but I'm pretty sure, by cross-checking, it's just a few hundreds of dB lower than displayed. I compensated for that.

I also compensated for temperature, since it's 29.5°C in my room by now.

Measuring the Millennia gain, which is part of the formula and therefore requires the same level of accuracy, is easier, since it's comparing the level with or without it. So you don't care of any RME error anymore, since it's removed at the end.

Any mistake here, in your opinion ?
 
Last edited:
Top Bottom