DIY Audio Projects Forum

Hello,

I am looking for answer of the thing which puzzle me, and interestingly I can't find it anywhere. It must be somewhere - hidden. The question is: how do I select crossover order for the driver(s), whether it is going to be first, second etc. order . Is it empirical process by that I mean: is it determined by driver's parameters; or, is it left at designer's will?

His "explanation" is kind of vague, at least to me.. Anybody would be willing to elaborate the statement in bold.

Thanks

Hi Aksa and welcome to the forum.

He is suggesting that trial and error and subjective listening session play a part in the crossover selection. I think that is true if you want to get the best sound. That may be practical in a commercial setting but not really for most DIYers.

I use 2nd and 4th order to keep the phase simple (2nd order most of the time). I'll use 4th order if the tweeter is crosses in low or if the woofer has too much higher frequency energy that you don't want coming through.

Cheers

_________________
[ DIY Mains AC Power Cable Cord ] - [ K-12 Tube Amp Kit ] - [ Tang Band D4-1 Horn Speaker Kit ] - [ Monoblock Push-Pull KT88 Tube Amp Kit ]

Thanks,

So, choice of crossover order has nothing to do with say: FR or impedance curve?

Yes, it does, particularly near the crossover point you are selecting. For example, you may want to XO at 2kHz, but the woofer has a nasty peak at 3kHz. Depending on high that peak reaches, you may prefer a 24dB slope (4th order) instead of a 12dB slope (2nd order) to suppress it.

The point being made in the LDC is that after driver measurements (don't trust the datasheet) and computer modelling to develop the crossover circuit, expect to fine tune it by trial and error to get the best sound. In short - it is not a trivial exercise. I usually suggest following proven diy designs.

Cheers

_________________
[ DIY Mains AC Power Cable Cord ] - [ K-12 Tube Amp Kit ] - [ Tang Band D4-1 Horn Speaker Kit ] - [ Monoblock Push-Pull KT88 Tube Amp Kit ]

". . . you may prefer a 24dB slope (4th order) instead of a 12dB slope (2nd order) to suppress it."

In other words, the one with sharp cutoff?

The second part of your comment has to do with my nature. Lynn Olson said that the most important part of the loudspeaker is the crossover. Very true, but how to do it? The devil is in details and venerable LDC is very poor and shallow in that regard.

Did you read Ray Alden's Speaker Building 201, is it better book?

And, Thank you very much.

No, I think the LDC is better. Here is a brief summary of the books:
http://diyaudioprojects.com/diy-audio/d ... -books.htm
Cheers

_________________
[ DIY Mains AC Power Cable Cord ] - [ K-12 Tube Amp Kit ] - [ Tang Band D4-1 Horn Speaker Kit ] - [ Monoblock Push-Pull KT88 Tube Amp Kit ]

Hi.

Indeed, very vague to me too. IMO, the author wanted to create the impression that the design of passive loudspeaker X'over networks, which sorta makes or breaks the sonics of a loudspeaker system, is something like rocket science, which greenhorns should not touch.

Back to your question, my straight forward answer is LESSer order of filter BETTER will be the sound. 1st-order filter (6 dB/octave down slope) will allow music signals to pass thru with minimum amplitude attenuation & least phase distortion. It is the phase distortion that screws up the music. Why? Less order fitlers needs less parts. Less reactive parts (capacitors & inductors) in the signal path therefore delay less on the music, thus better the music will sound. It is the most simple X-over design using least components, & is therefore cheaper to build.

But, 1st order passive X-over network is not all angel. It depends a big deal on the 'muscle' of the tweeter chosen as there will be larger overlapping of LF energy 'leaking' into the HF band which may damage the tweeter's fine voice cois. There's why we can see series resistor is often added upstream of a tweeter as current limiter.

Also it is least amplifier power efficient vs other orders (2nd-order=12dB/octave, 3rd-order=18dB/octave etc etc) as the larger overlapping LF/HF areas means more electrical power from the driving amplifier dissipated uselessly.

Also, always position loudspeaker with 1st-order fitler facing ON-axis to the listeners to minimize off-axis acoustic frequency response variation, the common weakness of lower orders of filters.

So my first choice of X-ver networks will be 1st order fitler which sounds best to my picky ears given the above consideration.

That done, the 2nd step is to make a passive X-over network more efficient & sound even better all around. (somewhat O.T.)

This is to modify the X-over network (be it 1st order or 4th order) to multi-pair-wiring instead of its common single-pair wiring. My years hands-on experience has proven
such modification works bigtime.

It improves the driving amp efficiency, raise the X-over power handling capacity & overall sound quality of the loudspeaker.

Hopefully I've pin-pointed the key knowledge one should know about X-over network design. Save studying those rocket-science-like loudspeaker "cookbooks".

c-J

_________________
"If the facts don't fit the theory, change the facts" quoted Albert Einstein.