Crossovers, minimalist vs Complex.

[Phototone]

I really can't explain it, as I am just using my ears, but it seems to me that some speaker systems that have absolutely minimal crossover components, things like just capacitors and resistors often sound just as good, or sometimes better (to my ears) than those speaker systems that have complex crossovers that have inductors, resistors and capacitors. It is curious. Of course I am referring to recapped units in each case. Is there any real reason for this?

 

[mashaffer]

Sure less funky phase shifting and more "even" impedance curves. That said my philosophy is make it as simple as possible but no simpler. A lot depends on the drivers. Some drivers absolutely need higher order filters to avoid some very ugly portions of their spectrum. If you have a woofer with a very smooth mid and high range then you can use either no LP or a simple inductor but if it has a nasty breakup in the mids then that won't work.

mike

 

[Coytee]

I once heard an Electrovoice Dx38 (active crossover with hundreds of do-hickies inside it) compared to a no expense spared, $3,000 in PARTS ALONE passive crossover on the same speaker.

I even saw the graph when they curved the speaker with both crossovers in their anechoic chamber. The graphs were indeed, very close but not exact.

I personally could NOT tell a beans worth of difference between the two crossovers. The owner/builder (not designer) of the $3,000 in parts cost alone passive, said he could hear more air or weight or bloom or something in female vocals....

I can't say he doesn't. I can only say that to my ears, he did a fantastic job of building this passive. I bought a Dx38 on Ebay for $400 and can get the same sound he paid $3,000 for parts alone.

For the difference in price, he can have the bloom that he hears and I do not!! I'll take the difference in cash & buy something else.

For what it's worth, as I understand their operation, they put the speaker in their anechoic chamber (this is Klipsch that I'm referring to) and tweak the sound using their Dx38. once they get the desired sound they then go about trying to recreate that same curve using passive parts.

Ultimately, I think (in their world) the rational way to present it would be that the passive is trying to recreate the sound they established by using an active. In the home, I'm using their same active, same speaker so in a sense, I'm getting closer to the sound they created in the chamber than the guy with the passive however, don't tell him.

I'd speculate it all comes down to there are good designs and bad designs. Ever seen an extreme slope passive designed by Al Klappenberger for a pair of Khorns?

www.alkeng.com

click on Klipsch and go to extreme slopes... these things are monsters but they sound very nice AND allow you to step up the volume and have the sound stay cleaner at bigger volumes.

 

[triode4]

I know that you need a separate amplifier for each driver for a crossover like the DX38. You only need one amp for a passive.So we need to add the expense of those other amplifiers.

 

[Zilch]

Many vintage designs manipulated driver parameters to perform the function, and minimalist filters were used primarily to "shape" the response with the drivers defining the fundamental character of the system. As the art evolved, more knowledge and better design tools permitted filter theory to assume more comprehensive control of the outcome.

 

[SaSi]

There was a time when a complex crossover was considered a prerequisite for a good loudspeaker system.

But "Occam's razor" suggests that the simplest solution to the crossover problem is usually the correct one.

 

[audiomagnate]

Dynaco A25s are a good example. One cap plus some resistors to adjust the tweeter level, IIRC.

 

[sealy]

Many vintage designs manipulated driver parameters to perform the function, and minimalist filters were used primarily to "shape" the response with the drivers defining the fundamental character of the sound. As the art evolved, more knowledge and better design tools permitted filter theory to assume more control of the outcome.

Genesis and EPI as examples. Woofers run full-range, with the tweeters designed to blend with the natural roll-off of the woofer. No inductors and just a single cap. Works wonderfully, with the Genesis especially. A friend and I each have a pair of large Genesis 210 speakers. We conducted several tests running the Genesis' against a pair of AR 9's, a pair of Fostex hybrids with planar mids and tweets and a huge Kenwood speaker system with bullet tweeters. Amps used were a monstrous Classe power amp, a Kenwood pure Class A integrated amp, an Audire power amp, a Conrad Johnson power amp and a British HH MOSFET power amp. Pre amp used was a Conrad Johnson. Listening tests took 1 hour, as we could not believe what we were hearing. Sonically, the tonal similarities of the Genesis' to the other much larger and much more expensive speakers was uncanny. Naturally the scale of sound and overall sound stage of the other speakers was larger, but sonically, the Genesis' more than held their own, including in the bass department. I doubt anyone has tested a pair of Genesis 210's to this degree, but in my opinion, these tests prove unequivocally that shaping driver parameters ,at least in this single instance, can achieve the same degree of success as the practice of employing very complex and costly crossovers.

 

[Phototone]

Glad to hear that I am not alone in feeling that a complex crossover is not a "requirement" for a good sounding system..it just depends on the design criteria. Those tube consoles that many people here seem to like don't have much of anything except a capacitor, ha ha.

 

[JoeESP9]

Electrostatics with sub woofers. No crossover from 80Hz up. If a simple minimal crossover is good none is even better.

 

[Phototone]

Electrostatics with sub woofers. No crossover from 80Hz up. If a simple minimal crossover is good none is even better.

Well sure, and single-driver "full range" speakers like Lowther don't have need for any either.

 

[PioneerGuy75]

Here is what I have noticed....
Simple crossovers = better matched drivers.

Complex crossovers = perhaps not as good matching between drivers?
Trying to make it work with multiple components?

Just my $0.02

 

[sealy]

Exactly right. With the advances in crossover technology, it is not as critical to have drivers that are matched precisely. Turns what used to be an art into a science.

 

[jamesm007]

I agree. Paradigm even stated 20 years ago that they design better drivers for less need of tweaking in the crossover. But Jim Thiel did not go to the same school as those guys, his are very elaborate to say the least. At one time he advertised crossovers that had more components than a preamp.

Here is one, its a first order two way cross-over for the CS6 speaker.

The technical reading of why

http://www.thielaudio.com/THIEL_Site05/PDF_files/PDF_tech_papers/techpaper_general1.pdf

and

http://images.google.com/imgres?img...ers&hl=en&rlz=1T4GGLL_enUS367US367&tbs=isch:1

Its amazing, he is correcting driver response, phase, off axis response, time response, and more...

I have had the opportunity to spend some hours with them and they are smooth and highly musical so I won't say it hurts the sound as it may be a physiological factor. The were highly coherent for a lack of a better word. But speakers with natural smooth drivers on and off axis, and with extended response like tweeters than can smoothly go to 1 khz makes for a more natural sound IMO. But not at Rock Concert volumes, the domes break up, and suffer power compression. Only negative that I can think of.

 

[68custom]

Here is what I have noticed....
Simple crossovers = better matched drivers.

Complex crossovers = perhaps not as good matching between drivers?
Trying to make it work with multiple components?

Just my $0.02

I would agree with this. sorta falls into KISS (keep - simple) way of thinking. but being able to design and match drivers and crossover is not an easy task. hats of to those able to build and design such a beast.

 

[whyaskit]

I want my crossovers to make the transition between drivers to be a smooth and pain free as possible. Whatever it takes and no more.

 

[sealy]

Totem Acoustics speakers are virtually crossover free. The Model 1 as a for instance, uses a single inductor, a series of resistors and that's it. The resistors are located outside of the cabinet as they generate a considerable amount of heat.

 

[JimPA]

Simple phase coherent is usually better.

 

[sealy]

Simple phase coherent is usually better.

By and large, I agree, although achieving phase coherence is far from simple. However, Totem's speakers are renowned for their accuracy, detail and sound stage depth. Not for their efficiency, though.

The crossovers in my Spicas are tailored for phase coherence and are also quite inefficient - 82-83 db/w, I believe.

 

[Michael Powers]

Two more come to mind....

Genesis and EPI as examples. Woofers run full-range, with the tweeters designed to blend with the natural roll-off of the woofer. No inductors and just a single cap. Works wonderfully, with the Genesis especially. ...

Some Epos models & Radio Shack's Minimus 25 were designed this way.