Tweeter Measurements
Copyright 2007 Troels Gravesen

The following few graphs and pics are meant to give a little advise regarding driver measurements. How is the microphone going to be mounted and what about the tweeter?

First of all:

The microphone:

I cannot enough recommend reading the articles written by John Atkinson at Stereophile:
http://www.stereophile.com/features/99


My microphone placed at the end of a 55 cm long rod. You simply need to do what's shown here above and below: Place the mic at the end of a long rod. 0.5 meter will usually do for 1 meter measurements. If you do measuments at 1.5 meter distance, you need a 0.75 meter rod.


Correct microphone set-up.


From this you will have severe reflections from the mic grip.


The problem is this. The soundwave from the speaker will hit the microphone grip and bounce back to the microphone capsule.


Found on the web: This may be good for room measurements, but not for speakers.


From the web: This definitely doesn't work.


Someone sent me this. Won't work either.

The good question is how smooth the rod + mic has to be not to reflect soundwaves back to the microphone capsule. Here's an example of how little it takes:


I've here placed a small piece of cardboard on the microphone rod. This is what it does:


Blue: Cardboard placed 12 cm from microphone capsule.


Blue: Cardboard placed 20 cm from microphone capsule.


Blue: Cardboard placed 30 cm from microphone capsule.

As can be seen this tiny reflector seriously compromises the tweeter performance.


Someone sent me this picture. Looks OK, but the tip of the long rod is a potential problem. I've tried adding a second 12 mm rod to my set-up and didn't record any disturbance. But take care. We can't fool acoustics.

Vifa XT25TG measurements


Above the Vifa XT25TG tweeter flush-mounted on a 40 x 50 cm baffle, a little off-set. Usually this tweeter has a smooth response from 1 kHz all the way to 20 kHz. And this reading is no exception. Measurement taken at metre distance with 3.07 ms window as seen on the impulse response below.


Impulse response.

Above is seen what we should not do. No flush mounting of the tweeter. Just cutting a hole for the magnet leaves a 4-5 mm step from the tweeter face plate to the baffle plane. The result is seen below.

Response of XT25TG from non-flush mounting. Quite a difference. Below the two readings on the same graph:

Blue = no flush mounting. Red = tweeter flush mounted.

This is what it should look like. Rebate a little deeper than the height of faceplate, add some rubber strips and screw the tweeter in place until the faceplate is flush with the baffle. No more, no less. Not 0.5 mm below and not 0.5 mm above. Flush!!

Make sure the rebate has the exact diameter of the faceplate + 0.1 mm. If wider you can use tape to cover the ditch. Even a narrow ditch will cause response irregularities.


Use thin painter's tape like seen above.