Hi LowQCab and thanks for the input.
These circuits are for pro audio gear with a balanced interconnection and 28V power. Also, they are quite complex. Not what I am looking for.
Thanks anyway, ab

Upon review of those circuits, it looks like the PU mic is a possible replacement for the 61A. Thanks.

 

A microphone has a very low signal level for nearby speech, therefore it needs a preamp that has a voltage gain from about 50 times to 200 times. your transistor had a voltage gain of only 10 times or less. Use an audio opamp.

This circuit and your explanation are invaluable to me. I am sketching up a board now. How about a 5532 for the op amp? Thanks.

A microphone has a very low signal level for nearby speech, therefore it needs a preamp that has a voltage gain from about 50 times to 200 times. your transistor had a voltage gain of only 10 times or less. Use an audio opamp.

 

A Dual- TI-5532 will work just fine.
Use both Amplifiers in series and set them up for ~10X-Adjustable-Gain each,
which will equal ~100X-total-Gain or less, if that's too much for your purposes.
.
.
.

 

The NE5532 is old and is a dual opamp.
The NE5534 is a better single opamp that is needed for a preamp.

 

This circuit and your explanation are invaluable to me. I am sketching up a board now. How about a 5532 for the op amp? Thanks.

You don't need a particularly good op-amp, because the signal-to-noise level of the microphone isn't very good - only about 70dB.
Because you are going to need large value resistors in the circuit, a JFET op-amp will be better than a bipolar, because its current noise is much lower so I would recommend the TL071 rather than the NE5532 (80fA/√Hz compared to 700fA/√Hz).
There are even better JFET op-amps but the microphone's self noise will dominate.

 

There are certainly better suited Op-Amps for critical Audio applications,
but for Speaker-Output-testing, the noise floor is largely irrelevant.
.
.
.

 

There are certainly better suited Op-Amps for critical Audio applications,
but for Speaker-Output-testing, the noise floor is largely irrelevant.
.
.
.

Depending on the sound level being measured, the amplifier noise level might matter. Certainly it needs to be considered before the results are to be trusted. Likewise any distortion products need to be understood , because they might have an effect.
The effect will vary with the actual sound levels and actual voltages.
This is not intended to be a challenge to any of the advice, but rather as a caution against unintended results.

 

We have no defined context as of yet.

Is knowing "how-high-is-up" actually necessary knowledge, or even useful at all to a hobbyist ?
Altitude above Sea-Level, and Amplifier-Output-Impedance,
can measurably affect a Speaker's Frequency-Response, do they matter ?, not to 99.99% of the population.

My "assumptions", ( which I always hate to have to make ),
always lean towards "Hobbyist-Grade" unless explicitly stated/requested otherwise.
Repeatable, coarse-differences in Frequency-Response are all that is really needed in this instance,
not "absolutes" with strict-definitions and tightly controlled conditions.

As far as Speaker-Design ........
A self-learning odyssey can certainly be personally rewarding,
but I'd rather learn from the documented experiences of others when
the journey is likely to get to the point of becoming life-consuming.
I wasted a lot of time teaching myself, the hard way, until I learned one particular thing .........
A Speaker is an "Energy-Transformer" with radically different Input and Output Impedances.
.
.
.

 

Assuming that you want 1V/Pa output (1Pa = 94dB)
The microphone has a current output so needs a transimpedance amplifier. It has a sensitivity of -35dBV/Pa with a 2.2k load, so you need a feedback resistor of 120k. Sensitivity is quoted at ±4dB so that would be 200k for a poor one and 75k for a good one.
There are five sources of noise:
1. the microphone itself, which has a SNR of 62dB. That's 800uV of noise at the output.
2. the op amp voltage noise. 15nV/√Hz for a TL071, which is multiplied by the "noise gain": the ratio of the feedback resistor to the bias resistor. Assuming a 10k bias resistor, the noise gain is 12. That gives a contribution of 25uV.
3. the op amp current noise (80fA/√Hz for a TL071), which develops a voltage across the bias resistor. and is multiplied by the gain. Contribution 1.3uV
4. The thermal noise of the bias resistor, 900nV, which is multiplied by the gain contributing 10uV
5. The thermal noise of the feedback resistor. 3.1uV
The noises are uncorrelated so add rms, total being 800.5uV

Note that swapping the TL071 for a NE5524 would reduce the op amp voltage noise by a factor of 3 to 8uV, but increase the current noise contribution by a factor of 5, also to about 8uV.

Also note that it would need an op-amp that is ten times worse to get anywhere near the microphone's self noise. We're into 741 or LM324 territory.

 

Some speakers use a piezo "shrieker" as a tweeter. Its frequency response has many narrow high frequency peaks and notches so it is difficult to measure and its output curve is usually smoothed.

 

We have no defined context as of yet.

Is knowing "how-high-is-up" actually necessary knowledge, or even useful at all to a hobbyist ?
Altitude above Sea-Level, and Amplifier-Output-Impedance,
can measurably affect a Speaker's Frequency-Response, do they matter ?, not to 99.99% of the population.

My "assumptions", ( which I always hate to have to make ),
always lean towards "Hobbyist-Grade" unless explicitly stated/requested otherwise.
Repeatable, coarse-differences in Frequency-Response are all that is really needed in this instance,
not "absolutes" with strict-definitions and tightly controlled conditions.

As far as Speaker-Design ........
A self-learning odyssey can certainly be personally rewarding,
but I'd rather learn from the documented experiences of others when
the journey is likely to get to the point of becoming life-consuming.
I wasted a lot of time teaching myself, the hard way, until I learned one particular thing .........
A Speaker is an "Energy-Transformer" with radically different Input and Output Impedances.
.
.
.

I have a simple way of avoiding the problem of "assumptions". You may have noticed it and not realized what it was. When needed information is not given, I state a "Guess", which is not as solid as an assumption. So there is a handy work-around for the problem.