"Budget" microphone measuring setup

[mrkva]

Hello everyone,

I am slowly setting up a lab to test my microphone products and am looking for some help with the setup from the local gurus  8)

What I measure (and want to in the future):
– sensitivity
– freq. response
– polar pattern
– THD

Here is what I gathered so far:
– RME Fireface UC soundcard
– Earthworks M30 reference microphone with calibration chart
– 5 digit bench DMM, 50MHz digital oscilloscope, lab power supply
– calibrated 94 dB 1kHz SPL pistophone generator with two adapters (1/2", 1/4")
– software: Spectre, Fuzzmeasure Pro, REW
– relatively well isolated room (former freezer) which I plan to make as anechoic as possible

I am looking for your advice with:
– software which would allow me to measure microphone's polar pattern
– software for automatised "production line" measurements (saving freq. responses and sensitivities to a database)
Mac, Linux or Windows, I can build a dedicated computer for this. I had some quotes but it is all in 10+k € region and that is way out of my budget – looking to invest couple of hundred € max

– reference speaker for mic measurements (lets stay sub 1000€)
I guess the best would be single "fullrange" driver or somewhat "co-centric" design with low harmonic distortion figures?

Thanks a lot for any suggestions! And please excuse my english, not a native speaker

 

[Bo Deadly]

I am looking for your advice with:
– software which would allow me to measure microphone's polar pattern
– software for automatised "production line" measurements (saving freq. responses and sensitivities to a database)
Mac, Linux or Windows, I can build a dedicated computer for this. I had some quotes but it is all in 10+k € region and that is way out of my budget – looking to invest couple of hundred € max

If you really want to get nerdy about it you could use a microcontroller / servo motor / position sensor connected by USB to a laptop running whatever software that can be automated. Use a few laser / plasma cut pieces to make a rotating "table" and you could polar pattern test all sorts of things.

 

[mrkva]

If you really want to get nerdy about it you could use a microcontroller / servo motor / position sensor connected by USB to a laptop running whatever software that can be automated. Use a few laser / plasma cut pieces to make a rotating "table" and you could polar pattern test all sorts of things.

Definitely planning that! Just built a custom coil winder with steppers the other day, so this should be much easier

 

[mrkva]

Very inspiring read from Earthworks: https://earthworksaudio.com/wp-content/uploads/2012/07/how-earthworks-measures-mics.pdf

 

[joaquins]

I think you are in the right track, starting with a standard microphone and a calibrator is crucial I guess.

...
– relatively well isolated room (former freezer) which I plan to make as anechoic as possible
...

  I want to comment about this in particular. Having a small anechoic space is quite hard, specially for low frequency, if not impossible. What's the frequency range you are interested in? For what measurements? You might not want to do them all in the same space.
  Good news is you don't really need the space to be anechoic, but you do need to be big. If make your measurements in a BIG room, you can forget about reflections, as you only use the measurement of your firsts few ms. How much ms depends on the size of the room, but basically what you want is to have your first reflection outside of the window, you put a short burst of reference signal, capture that window that has no reflections and do math with it, repeat till you get enough data to make up the short measurements for the precision you need.
  That being said, you want your closest wall to be ~9m away to measure 20Hz so you can get a full cycle out of your measurements before the first reflection comes back, with that your FFT will have 20Hz steps. There are walkarounds so you don't need 20m height room, like burring the speaker in a corner so it's first reflections only come from the oposite corners but might be a problem with high frequency, hence you might not want to share the same space for different frequencies.
  Other used technique to run acoustic meassurements is inside a tube, I've seen it applied to meassure absorbent materials but you could apply it to microphones, 10m long PVC tube looks more plausible and cheap than 10m or 20m sided room. Placing the speaker at one side and a mic 1m away gives you enough time to measure before the first reflection to come back.

BTW, I love fuzz measure, nice pice of software, haven't used in a while but I recall it to be great.

JS