1st order filters/crossovers with proper drivers compared to full range?

[tech.knockout]

Well assuming I have very well behaved drivers that are up to the task, what disadvantages do a 1st order low pass on a woofer have compared to full range? (when crossing at around 250hz - the high pass will be the natural roll off of the midrange)


1st order filters have no/negligable phase problems. (correct me if im wrong here - as i heard 250hz is in a phase sensitive region?)

They have very minimal componentry, so distortion caused by componentry is negligable as well.

[Michail Bugaria]

Intersting article
Audibility of different crossover types

[inertial]

Hi,
yes,very interesting indeed.
IMHO I am not shure 300-8000Hz is "sufficient".
Maybe a pair of Hi-Vi B3 "a la Zaph" can be more correct.
Again the man do not explain very well the room setup and talking about "normal living room". I think a very well control of first reflections are obvious and a little listen triangle also (1,5-2mt) but
he do not specificate.
Anyway seems to me a serious work.

cheers,
Inertial

[cotdt]

yeah i think all the audibility comes from the differing power response of different order crossovers.

[LineSource]

I personally use very steep filters like LR8 in order to put 80-3,000Hz human voice on the midrange driver of a 3-way.

I will move to steeper digital filters plus digital room equalization and run a digital coax to each speaker when the DIY parts are available.


John Kreskovsky is a well regarded expert who has studied crossover design from the power delivery aspect.

Consideration of Power Response in Speaker Design: Crossover Choice and Setup


http://www.musicanddesign.com/Power.html

http://www.musicanddesign.com/

[ScottG]

Originally posted by tech.knockout
Well assuming I have very well behaved drivers that are up to the task, what disadvantages do a 1st order low pass on a woofer have compared to full range? (when crossing at around 250hz - the high pass will be the natural roll off of the midrange)


1st order filters have no/negligable phase problems. (correct me if im wrong here - as i heard 250hz is in a phase sensitive region?)

They have very minimal componentry, so distortion caused by componentry is negligable as well.

In this case you need to consider the effects of the inductor to the sound. Particularly hysterisis of the core if not air-cored, OR if air cored how much additional amp current will be "burnt-up" because of resistance (..and what it will do to amplifier dampening).

You also need to consider the woofer's contribution. Notably the passband above the lowpass where driver break-up could substantially impact the sound. Furthermore you need to consider the loss of point-source character. Finally driver mass and compliance should be moderatly similar between the two drivers.

Still, I personally think there is more to be "gained" here (if done properly), than "lost".

Here is a listing for inductor values for 8 and 4 ohm drivers:

http://www.colomar.com/Shavano/crossover6db.html

If you can afford the initial expense of the wire AND can test for inductance values, then you will likely be better served by making your own inductors for anything above 2 mH.

[Curmudgeon]

Yes, a very interesting article. It is a nice starting point. However in actual designs, the crossover is also used to equalize; programs such as LEAP can optimize for on-axis response or total power, or whatever. An obvious modification to most of the "pure" filters would be to slightly overlap the crossover frequencies giving the effect seen with the Bessel in the article. This is the sort of thing that is done during final voicing.

To return to the original question; no full range can equal a well designed multi-driver system for frequency range and smoothness. Some listeners feel that full range drivers have a greater sense of immediacy; others hear the immediacy as the result of slight peakiness. Mileages vary; it might be fun to try both.

The quality of the drivers will have a very large impact on the results; you will always wind up at best comparing apples and oranges, or perhaps bananas and blueberries. Closest might be to use a full range driver, and compare it to a system that uses that driver as a midrange; they usually serve as excellent (albeit pricey) midranges.

Also note that a crossover using substandard components is just going to muddy the waters, and the music; good components are a must for a fair comparison.

[tech.knockout]

Originally posted by ScottG
Furthermore you need to consider the loss of point-source character. .

Yes the point source seems to be the greatest worry. What specifically are the problems of lacking point source character (or a signal being split between two drivers)?

But then line arrays(the ultimate for most people) arent point sources.

Finally driver mass and compliance should be moderatly similar between the two drivers

Would the problem of drivers having their own unique character be minimized if the drivers have very low distortion/good transient response in the first place?


Power response is basically frequency response, is it not? If so the full range would be inferior here anyway.

[cotdt]

if you want point source you can use coaxials. personally i don't think there is much to benefit from, since in my 2-way speaker i can't even tell my tweeter is there. it certainly sounds like a point source to me, although it isn't.