Basic measurement question

[dynamo]

When measuring drivers in baffle near field, is there a certain distance you shouldn't be closer than to be all inclusive of baffle effects like bsc and edge diffraction?

Thanks in advance

[dcibel]

Re: Basic measurement question

Part of the point of near field measurement is to not have baffle effects / edge diffraction present, as well as any room reflections. 1/4" is good, you want to be as close as possible to have just the driver output and nothing else. You then apply the BSC simulation and merge with the far-field measurement to get full frequency response.

[dynamo]

Re: Basic measurement question

Sounds good, thank you

[jonpike]

Re: Basic measurement question

When measuring drivers in baffle near field, is there a certain distance you shouldn't be closer than to be all inclusive of baffle effects like bsc and edge diffraction?

Thanks in advance

Yep, it's more a distance you don't want to be farther than. The rule of thumb formula is closer than 1/10th the radius of the drivers cone, in the center. So a 6.5" driver might have say a 5" cone, a 2.5" radius, and thereby .25" spacing or less for a good NF. The only way to be too close, is to be so close the cone bangs the mic when you apply the test signal! Then move out a bit, or lower the level some.

[Jeff B.]

Re: Basic measurement question

Conversely, you are effectively in the far-field when you are at a distance that is more than 3 times the radiating diameter of the driver, and double that of the baffle width. So for most small two-way speakers placing your mic 18 - 20" will produce good results. You don't have to be at one meter, and getting farther away just makes it harder to get room effects out of the data. Here's a link to a paper I handed out at my seminar on taking measurements.

https://app.box.com/s/s564qmm526fuqlym0eez

[dynamo]

Re: Basic measurement question

Thank you both, and thanks for the link to your paper Jeff-very helpful as always. The reason I ask is that complex baffles aren't really always able to be modeled. Thanks much guys.

[300Z]

Re: Basic measurement question

Conversely, you are effectively in the far-field when you are at a distance that is more than 3 times the radiating diameter of the driver, and double that of the baffle width. So for most small two-way speakers placing your mic 18 - 20" will produce good results. You don't have to be at one meter, and getting farther away just makes it harder to get room effects out of the data. Here's a link to a paper I handed out at my seminar on taking measurements.

https://app.box.com/s/s564qmm526fuqlym0eez

Thank you for sharing the paper Jeff!

[DDF]

Re: Basic measurement question

I modeled the difference with distance, attached. About half a dB, assuming the PE .75 cubic foot box, and mic in front of a standard 6.5 inch driver

Not much, but something to definitely consider in design as even half a dB is audible over such a wide bandwidth

[Jeff B.]

Re: Basic measurement question

I modeled the difference with distance, attached. About half a dB, assuming the PE .75 cubic foot box, and mic in front of a standard 6.5 inch driver

Not much, but something to definitely consider in design as even half a dB is audible over such a wide bandwidth

You're right, Dave. But for people measuring indoors the closer distance allows for longer gate times before the room effects begin to take over, so closer often nets you some lower frequency resolution. As with any acoustic measuring - trade-offs abound.

[DDF]

Re: Basic measurement question

You're right, Dave. But for people measuring indoors the closer distance allows for longer gate times before the room effects begin to take over, so closer often nets you some lower frequency resolution. As with any acoustic measuring - trade-offs abound.

No need for trade offs, measure twice. Longer distance for BDC effect, closer if you want higher res for any driver break up effects.

Most people use FFT window types that inherently smooth the result. Using longer distances and rectangular windows (carefully selecting window so that time record and its derivatives are as close to zero as possible at the window boundary) provide a good trade off between resolution and accuracy.