Best Practices for distortion measurements on speakers?

[sondek12]

Looking for best practices on how to perform distortion measurements on loudspeakers.

How do I minimize the influence from room modes and other rattles and buzzes when I try to measure in a normal non anechoic room? Any specific properties, limitations or requirements on the room or set up?

How do I minimize the influence from background noise and wind when measuring outside?

Measurement system:
- ARTA STEPS 1.7.0 software
- Windows XP on a Powerbook Pro
- Earthworks M30BX microphone
- EMU USB Tracker Pre external soundcard

Looking forward to all tips and experiences.

Best Regards, Mats

[audiomik]

You should find this Application Note from Brüel & Kjær of interest as it covers both theory and some methods for measurements:
http://www.bksv.com/doc/bo0385.pdf

Also this info page from Audio Precision, explains why it is extremely difficult to take even basic distortion measurements without either good anechoic, outdoor conditions or measurement techniques capable of substantial environmental compensation:
http://ap.com/solutions/loudspeakers

Not knowing the capabilities of your specific measurement hardware/software, I somehow doubt it is capable of the compensation needed if AP can't achieve this with their products!

anyway, a starting point for your question......
Mik

[sondek12]

Thanks for the tips on the articles. I have read the B&K one, which gave me some insight. I understand the challenges you have when you try to measure in an non anechoic environment. I just haven't found out just how the ARTA system can help me to overcome these limitations.

I have tried to use the STEPS functionality, but have so far failed to set it up to gate the measurement in order to minimize room effects. I will now explore the chirp function in ARTA instead, and see if I can get more reliable results from that.

Please keep forwarding your best practices. I want to learn how to overcome the challenges to measure distortion on my speakers.

Regards, Mats

[classicaudio]

Here's the cheap and dirty.

On a wind and rain-free day, Haul the speakers into the backyard. Set on the grass with a towel beneath at some distance from any reflective boundary. Run a sinewave into the speaker and place a condenser mic a meter away from the driver that will be handling the lion's share of the test frequency. Use a mic stand.The mic should run into a cassette deck or reel-to-reel with a decent mic amplifier, and the output should run into a distortion analyzer.

Then wait for one of those random lulls in background noise, and take your measurement.