MEasurement distance

Hi everyone,

I've done some measurements on drivers in enclosure.
According to WinISD, each driver has plenty of air-volume behind it.
Still, every driver except the tweeter, rolls off way to early in the low end, compared to the data sheets.

Measurement distance is 50 cm. Can that be the cause? If so, what's the math?

Jennice
 

Svante

Member
2004-02-01 11:17 pm
Stockholm
Jennice said:
Hi everyone,

I've done some measurements on drivers in enclosure.
According to WinISD, each driver has plenty of air-volume behind it.
Still, every driver except the tweeter, rolls off way to early in the low end, compared to the data sheets.

Measurement distance is 50 cm. Can that be the cause? If so, what's the math?

Jennice

Could you post the measurements and simulations?
 
The attached image shows the frequency/SPL plot for the HDS-134 up to 1 kHz.

It's mounted in a box of 8.5 liters ( - some fill material on all sides, except baffle).

Therefore, the device shouldn't drop off this early?!

I am thinking if there is a relationship between the measured frequency's wavelength and the distance to the mic.?

Jennice
 

Attachments

  • hds-134.gif
    hds-134.gif
    13.3 KB · Views: 185

Svante

Member
2004-02-01 11:17 pm
Stockholm
Yes, it seems that you have a 6 dB baffle "step" (higher level for higher frequencies), and that this step is somewhat hard to see because of the room interference that you have in your measurement (you measured in a room with a sweep, right)?

The microphone distance will not show up in your measurement at a specific frequency directly related to the distance/wavelength. It will however affect the response curve. If you put the microphone closer, the baffle step will be smaller, since the distance to the diffracting edges becomes larger than the distance to the driver. The room reflections/resonances will be smaller in amplitude and the curve will look smoother. At greater distance, the baffle step will be fully visible and room resonances too.

If you want to experiment with the baffle step you can try The Edge in my signature.
 
AAaarrghhh...!!! :headbash: :headbash: :headbash: :headbash: :headbash: :headbash:

Yes, I have "the edge" (nice tool, BTW!). I guess I'm just plain stupid not to have considered this when looking at the data. :clown:

Maybe I can excuse myself, having the flu at the moment. I have a 28 cm wide baffle, which should fit nicely.

Interesting measurement ideas, though. If I understand you correctly, moving the microphone closer to the driver will (wich makes sense to me) make the diffused sound reflections less significant.

If I understand you correctly, measuring close to the driver will make a sort of pressure field measurement with less influence of the baffle step. I should be VERY close to the driver to avoid measuring the baffle step "losses", right? How close is OK?

Jennice
 

Svante

Member
2004-02-01 11:17 pm
Stockholm
Jennice said:
AAaarrghhh...!!! :headbash: :headbash: :headbash: :headbash: :headbash: :headbash:


:D

Originally posted by Jennice If I understand you correctly, measuring close to the driver will make a sort of pressure field measurement with less influence of the baffle step. I should be VERY close to the driver to avoid measuring the baffle step "losses", right? How close is OK?

Jennice

As always "it depends". But if we are talking low frequencies there is no problem at all, as long as the mic doesn't hit the membrane. Absolute level calibration will be difficult, but again that is mostly not important.

There will be an error towards higher frequencies, when the membrane does not act as a rigid piston, since the parts of the membrane that is close to the mic will be given greater weight.

May I also suggest a method to measure the response inside the box, you will need to tilt the response by 12 dB/oct (boost high frequencies) in order to translate the curve to the outside response. This works up to the first standing wave in the box.

You are danish, right? If you can read swedish, you might want to have a look at the loudspeaker lab in a course I give at KTH in Stockholm. The measurement method is described there.
Loudspeaker lab
 
Hi all,

I have made new meadurements, with the microphone much closer the the driver. Now the baffle step is almost invisible, and the curve nicer. The phase still wraps around 2 times. Where / how do I enter the compensation (speaker-to-mic distance) to "linearize" the phase?

Hope someone knows -
Jennice