Speaker Workshop: tweeter's built-in filter

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
This is from Speaker Workshop:

"Tweeters are really just drivers with built-in enclosures. They come in 2nd order and 4th order (most often) depending on whether or not they are a vented design. A 2nd order tweeter melded with the right 2nd order crossover can produce a good 4th order acoustic crossover. When using a 4th order tweeter use a higher order crossover to improve power handling."

How do you know that they are 4th order or 2nd order tweeter? And how can this relate to a vented (box)? Anyone care to explain? I'm also confused with the last statement

:confused:
 
diyAudio Member
Joined 2004
The rate of the natural acoustic rolloff of the tweeter is 2nd or 4th order, determined by the mechanical properties of the tweeter.

Then when you apply a 2nd order XO you get 4th order etc.

Virtually all tweeters have no rear wave so loading in a cabinet makes no difference. Baffle step does make differences though.
 
Oh, thank you all. I see it now...

So the rate of the order can simply be seen from the relative steepness of the natural roll-off graph/pattern. Whatever causes that! Initially I thought the 'vented' design has something to do with tweeter chambers, because what has it to do with speaker enclosure?

So the statement is referring to crossing the tweeter at it's natural roll-off to create higher order natural/accoustic filter, and because it is close to Fs (especially for 4th order natural roll-off?????) then high order filter is mandatory to improve power handling.

:cool: (I did that intentionally only with woofer, because it doesn't have the Fs issue)

DSP_Geek, some tweeters like Dynaudio D21 like to be crossed near Fs. Only for low SPL and that's the headache indeed.
 
Jay said:
Oh, thank you all. I see it now...

DSP_Geek, some tweeters like Dynaudio D21 like to be crossed near Fs. Only for low SPL and that's the headache indeed.

Mmm. Note that crossing a driver near Fs, unless the impedance peak is compensated with a conjugate network, means the response around there might not be what you expect. Perhaps that non-flatness compensated for a failing elsewhere in the system, maybe the woofer had a dip at that point.

I'm not saying you're wrong, just that I don't know the variables about the time you tried it. I still prefer to cross my tweeters at least an octave above Fs.

Cone mids, on the other hand, are entirely fair game for this approach.


Francois.
 
Acoustically a closed back driver is a second order high pass filter, and if it has a q=.707, cascading it with a Butterworth filter of the same frequency gives an exact L-R type chracteristic. Crossing it over at any other frequency will not give this due to residual phase shift, unless a Linkwitz transform "bi-quad" filter is used.
There also exist "quasi L-R" crossovers that allow crossovers in the f3-2f3 region of the tweeter, whilst still retaining the in phase through the crossover region property of the L-R.
 
Cal Weldon said:
I've not heard of that before.

Actually I didn't mean that crossing D21 near Fs is good. What I mean is that in many tweeters, they tend to sound better when crossed low (assumming that the resonance should have been taken care of). This makes tweeters to sound open. And crossing low means crossing near Fs, doesn't it? Besides, D21 has Fs at 1K3! (This gives subjective truth, since not many woofers can have good midrange above 3K)

XT25 is the best exception. Initially I thought the Quality/Price will be lower to cross it at high Fc, actually I found the opposite tends to be true. But I still have some doubt about the 'true' x-over point of XT25 crossover. If you 'cross' XT25 and another tweeter at the same frequency and order, the XT25 will outputs more vocal (and is very sweet :D, but not my cup of tea ;)) as if it has shallower roll-off, and it seems it still takes the benefit of low Fs.
 
DSP_Geek said:

I'm not saying you're wrong, just that I don't know the variables about the time you tried it. I still prefer to cross my tweeters at least an octave above Fs.
Francois.

Of-course I can be wrong. But I tried it in many occasions. I used to have inductors in almost every values. I tried many crossovers for every pairs of drivers.

With Fs at 1K3, what F will you cross the tweeter at? I have had plenty of headaches crossing woofers at 3K :bawling:
 
rcw said:
Acoustically a closed back driver is a second order high pass filter, and if it has a q=.707, cascading it with a Butterworth filter of the same frequency gives an exact L-R type chracteristic. Crossing it over at any other frequency will not give this due to residual phase shift, unless a Linkwitz transform "bi-quad" filter is used.

So at exactly what frequency it should be crossed at (to get L-R characteristics)? Right at -3dB?
 
re crossover

Yes Jay the crossover frequency should be the drivers resonant frequency. High quality tweeters sometimes have a qt that is too low and this can be modified with a series resistor, you can also lower qt, but this is more complicated.
In the case where the tweeter power handling is marginal at this low crossover frequency things can be improved with the Thiele fourth plus fifth order network.
 
It will be interesting to try it out, but with 4th order filter I have never got any tweeters to sound open/natural!

I intend to build a bookshelf using P13. In the old day, I examined D27 and D25, and I chose D25 to go with P13 (I liked D25 better than D27). But now I have D27 at hand so I experimented with many 4th order crossovers for the D27 from 2K2 to 3K, and I don't think good result is possible. Will try 3rd order, because I think it is the minimum order I should use in a small bookshelf.

I have a quite enjoyable bookshelf using D27. The strange thing is, the tweeter is 2nd order and the woofer is 3rd order. :cannotbe:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.