Active filters better compensate for FR anomalies?

[rick57]

What is the easiest way to *properly fix multiple peaks/ cone break ups/ harshness, without causing other problems?

A friend with expensive Focal 7K2 Kevlar drivers that sound quite shouty/ harsh, had at different times, three people experienced with passive XOs (one has done about 60 ) try to smooth them passively. When they got rid of the harshness with notch filters ~ apparently the dynamics were lost.

Linkwitz said in the Phoenix FAQ (www.linkwitzlab.com/faq.htm#Q26) :

“An understanding of the transfer functions of drivers and cabinets . . can lead to the design of active networks that *precisely compensate for frequency response variations and give the desired overall response in magnitude and phase.

Since opamps separate and buffer each filter section the design and fine tuning of the circuit can proceed without interaction.
Modern opamps, such as the OPA2134 are sonically transparent, and absolutely no cause for concern

Are active filters really more *precise?

With *multiple peaks/ cone break ups/ harshness, is harshness easier to cure with *active notch filters?

Thanks

[Sheldon]

It is possible to avoid breakup issues with passive or active filters. I have one set of speakers with Seas Excell drivers, the same that the Orion uses. The crossover is passive, but obviously well designed, because there is no hint of harshness or breakup.

But your question was "what's the easiest?" Tough question, but I would say they are about the same. I would think that active filtering will make it easier to maintain dynamics. To do either properly, you'd have to get a good measuring/design system. I think many of them will model both passive and active filters. Some will also allow you to test soundcard versions of the designed filters on your speakers. Active filters are more flexible and you have more choices, and it's easier to make higher order filters, but you have to build more circuitry and have more amps.

I took the easiest way out and have a DEQX, which allows me to, measure and make the filters and implement them, all in one device. Aside from the extra amps, it is in my view the easiest approach. I think the combined Berhinger crossover and equalizer can do much the same for much less, though without the linear phase filters and driver phase matching features of the DEQX.

Sheldon

[PMA]

Definitely - every passive crossover is only an egg-dance. You have to design complex circuits to supress resonant response, that influences filter response below/above roloff freq. Filter response can be done much accurately by active filters and you can compensate a lot of driver inherent problems. And resonant behavior is damped by amplifier itself. I recommend to read Siegfried's webpages, excellent stuff.

[soongsc]

I remember some threads talking about modification of TB drivers. FR ripples associated with phase ripple indicates cone breakup. If approached using some of these approaches, the results can be stunning. I beleive Kevlar cones can tamed by using things as simple as Colgate toothpaste if applied to the right places.

[rick57]

Hi

> To do either properly, you'd have to get a good measuring/ design system.

Yes, I not long ago acquired SoundEasy.
Though I have no idea if that will make it easier/ better with passive or active.

> Active filters are more flexible and you have more choices, and it's easier to make higher order filters, but you have to build more circuitry and have more amps.

Ok, and notch filters?

I was intending to get the Behringer DCX2496, I’ll probably get its replacement.

Cheers

[rick57]

Hi PMA

I was thinking that way

> Filter response can be done much accurately by active filters and you can compensate a lot of driver inherent problems

can you elaborate on that/ give an example?

Cheers

[Sheldon]

Originally posted by PMA
You have to design complex circuits to supress resonant response, that influences filter response below/above roloff freq.

Very good point. So all things considered, it probably is in the end easier to tame these issues with active filters - even accounting for the extra circuitry. So I guess that I should have said that neither approach - active or passive - is "easy", unless you use an completely integrated system like the DEQX (no affiliation, etc., etc.,). Same old choice. You can pick two out of three - easy, cheap, good.

Sheldon

[rick57]

Hi soongsc

You could well be right, but the science of toothpaste may not be as precise as the science of filters

Cheers

[rick57]

Hi Sheldon

> all things considered, it probably is in the end easier to tame these issues with active filters - even accounting for the extra circuitry.

Thanks, hunch appears to be confirmed . .


Cheers

[Sheldon]

Originally posted by rick57
Hi PMA

I was thinking that way

> Filter response can be done much accurately by active filters and you can compensate a lot of driver inherent problems

can you elaborate on that/ give an example?

Cheers

He did indirectly. Linkwitz's crossovers for the Seas drivers I believe incorporate notch filters. But in any case, he gives the design principles.

With steep FIR filters (as used with the DEQX or some PC based systems), there is no need for notch filters, as you can completely avoid the breakup region.

Sheldon