Home Theater Forum and Systems banner

1 - 9 of 9 Posts

·
Registered
Joined
·
168 Posts
Is it possible to use measure with REW,
in order to find what the crossover frequencies of a Passive 3-Way speaker are?
Thanks
REW has the required measurement capabilities but the problem is to be able to play each individual drivers at the same time. Do you have 3 pairs of binding posts at the back? Otherwise it may still be feasible but more complicated. Please report back and we'll see.
 

·
Registered
Joined
·
32 Posts
Discussion Starter #6
SPL in REW was calibrated @ 90dB was set from the 8" woofer using Pink Noise generator in REW.
No level changes or recalibration from speaker to speaker, same level kept.
Using ECM8000 (Waiting on UMIK-1)
1" from each cone. Woofer, Mid, Tweet
Attempted using Armacell 1"thick closed cell foam to isolate each cone
from the adjacent, not much change in measurement/no interference from
adjacent speaker.
On side for easy measurement without mic boom height change.
Regardless of smoothing, x-over points are at same locations.
What can you make from this?

Even if I fully cover the midrange with Armacell closed cell foam,
flat against box, same trace on tweeter.


 

·
Super Moderator
Joined
·
9,252 Posts

The absolute best way to measure would be to disconnect two of the drivers at a time.

That said, assuming the blue trace is the midrange, it’s curious and unusual that the midrange and tweeter overlap across a one-octave range between 4-8 kHz. The woofer low pass appears to have ~12 dB/octave slope, while the tweeter and mid appear to have ~18 dB/octave slopes.

Regards,
Wayne

 

·
Registered
Joined
·
168 Posts
The absolute best way to measure would be to disconnect two of the drivers at a time.
+1

That said, assuming the blue trace is the midrange, it’s curious and unusual that the midrange and tweeter overlap across a one-octave range between 4-8 kHz.
Agree that the blue curve should logically be the midrange but you also get a lot of the tweeter’s response in the blue curve since the tweeter is omnidirectional at these frequencies (ka < 2 at 4kHz). This is likely why both curves are parallel in that range, too. All levels being equal, the upper frequency driver will always “leak” more into the lower frequency driver’s measurement than the opposite (since the larger driver starts to “beam” at some frequency). My guess is the crossover is minimal.

In the OP's curves you see a lot these summations - one way to spot them is by looking at the differences between curves to see if they are correlated (if you see two parallel curves with the same ripples they are likely interactions). The unknown (relative) phase and delay makes it tricky… maybe the phase would be telling something.
 
1 - 9 of 9 Posts
Top