[kyrill]

maybe some ppl who still "defend" passive in box crossovers
should listen longer or more concentrated to good active solutions
I have a DEQX not a DEQ

the DEQX can actively surpass all passive high end Xovers  and the differences heard
underlines JLM post

Is it traditional thinking, "culturally" bound which keep speaker buyers and manufactures together with in box Xovers?
On the other hand how can a speaker manufacturer bypass passive Xovers?
what is left is a box and single drivers unless
designing an "audiophile" active Xover according to his taste..

Ah well..

[Scott F.]

(big snip)........ they do not magically solve the problems of designing a quality loudspeaker using real-world drivers. 

Guys,

I guess that is my real issue with all of this. The drivers you speaker designers have to choose from can be less than design friendly (in my opinion). Then again I were King of the Speaker World, ALL speakers produced could be easily driven to ear splitting levels with 2.5 watts on a zero feedback amplifier, produce deep bass in a 2cuft cabinet, have an even impedance curve without weird phase anomalies, rolled off at a nice smooth -6db per octave and showed no signs of cone breakup and was uber linear with little or no nasty distortion characteristics. Better yet if this driver from dreamland not only did bass but extended at least up to 14k. Thats not asking too much....is it? Unfortunately this isn't the case. You guys are forced to drive finish nails with sledgehammers at times.....and believe me, I DO understand exactly what you are talking about when it comes to XO design, I didn't just fall off the turnip truck.

I guess my issue is more with the industry that continues to make mediocre drivers (that should precipitate some nifty responses) and forces your (collective) hands into to using more and more complex filters in speaker designs. Thats not to say you guys don't push these speakers and designs to get the last ounce of music out of them but you have to admit that if you could get by with fewer crossover parts and shallower slopes you'd jump at the chance....wouldn't you?

As I said before, get rid of your passive crossovers and free your speakers. Until you hear what the advantages are, you will never know what you are missing. As for burden of proof, Ryan lets get together sometime later this summer and play with a few different ideas. Maybe pick a driver or two that you've got in stock that rolls off smoothly (ie no heavy cone breakup) where we can bi/tri amp the conceptual design using an active XO. If you want, we can use my Focal's, they roll off reasonably smooth. We can knock together some cabinets up in the wood shop. You can put together a quick passive XO and we can listen to them both ways. Then we can let our ears decide which is better. I think you will be surprised.

As I said in my first post, everything in audio is a compromise. On the speaker design side, more complex XO's are just that, a compromise you use to control drivers and their responses. You play the cards you are dealt. It just doesn't fit into my Utopian view of what audio should be. Then again, I seem to 'buck' the norm more often than not.

[jon_010101]

I think one of the biggest advantages of line-level crossovers, which I haven't seen mentioned here, has much to do with the amplification.  If you divide the signal at ~2kHz, all of the high order (and most 2nd order) harmonic distortion near the crossover point, and right near the key range for vocals, never make it to the tweeter.  The woofer can barely reproduce a 4kHz, or 6kHz, or 8kHz harmonic.  Not hearing that slightly-phase-shifted annoying distortion makes a tremendous difference, and takes some burden off of the amplifier.  By dividing the signal three ways actively, you probably can dramatically improve system THD -- woofer amp distortion never makes it to the midrange, midrange amp distortion never makes it to the tweeter.  Sweet!   

Now if only I could afford a three-way, tri-amped, active high-order three-way crossed-over, system -- and then fit it into my 12x12 room 

[cryoparts]

Also, note that effective implementation of an active crossover, analog or digital, is not trivial.  Active crossovers are a tool, and the results will be no better than the capabilities of the man using the tool.  They are not a panacea - while active (especially DSP) crossovers can often address issues that are not readily addressed with a passive circuit, they do not magically solve the problems of designing a quality loudspeaker using real-world drivers. 

Duke

Beautifully said, Duke.

Peace,

Lee

[cryoparts]

While the finest system I heard at RMAF 2007 and CES 2008 (the Sonicweld Pulserod) used a hotrodded DSP crossover, achieving that system's level of transparency is not trivial and so I am not sure that it would be cost-effective in most situations.

If one only knew the dedication to his craft, time spent perfecting said craft, and in-depth digital and DSP design knowledge that Josh Heiner (Sonicweld) has, they would quickly realize that utilizing an active crossover with DSP it is not a trivial undertaking or cost-effective solution whatsoever.  However, to my ears, his speakers produce the finest reproduction of recorded music I've ever heard.

He is slammed building product, doing high end medical and military design projects, so I doubt seriously he would ever find the time to post here, but visit our room at RMAF and feel free to ask him how much work and how many years he has invested to be at the point he is, you'll get an earful!   

Disclaimer--Josh is my closest friend and I have watched his growth as a designer over the last 15+ years, so call me biased if you'd like. 

Peace,

Lee     

[Vapor Audio]

I guess that is my real issue with all of this. The drivers you speaker designers have to choose from can be less than design friendly (in my opinion). Then again I were King of the Speaker World, ALL speakers produced could be easily driven to ear splitting levels with 2.5 watts on a zero feedback amplifier, produce deep bass in a 2cuft cabinet, have an even impedance curve without weird phase anomalies, rolled off at a nice smooth -6db per octave and showed no signs of cone breakup and was uber linear with little or no nasty distortion characteristics. Better yet if this driver from dreamland not only did bass but extended at least up to 14k. Thats not asking too much....is it? Unfortunately this isn't the case. You guys are forced to drive finish nails with sledgehammers at times.....and believe me, I DO understand exactly what you are talking about when it comes to XO design, I didn't just fall off the turnip truck.

I guess my issue is more with the industry that continues to make mediocre drivers (that should precipitate some nifty responses) and forces your (collective) hands into to using more and more complex filters in speaker designs. Thats not to say you guys don't push these speakers and designs to get the last ounce of music out of them but you have to admit that if you could get by with fewer crossover parts and shallower slopes you'd jump at the chance....wouldn't you?

I agree for the most part, but also understand completely why drivers are the way they are right now.  If there was Exxon money in making drivers the materials science research and application testing would be to the point where we could have a pistonic driver with no break-up at all, and motor designs far exceeding what we have today.  But given where things are, I think we have some pretty killer driver choices.  Hey, these RAAL ribbons are ruler flat and give a 24db acoustic rolloff with just a single cap 

But yes, definately ... if I could have my Seas Mag cone sound and detail, but throw in zero inductive rise and a natural 12db rolloff at 2500hz, and pistonic behavior to beyond 20K, you better believe I'd take it!

As I said before, get rid of your passive crossovers and free your speakers. Until you hear what the advantages are, you will never know what you are missing. As for burden of proof, Ryan lets get together sometime later this summer and play with a few different ideas. Maybe pick a driver or two that you've got in stock that rolls off smoothly (ie no heavy cone breakup) where we can bi/tri amp the conceptual design using an active XO. If you want, we can use my Focal's, they roll off reasonably smooth. We can knock together some cabinets up in the wood shop. You can put together a quick passive XO and we can listen to them both ways. Then we can let our ears decide which is better. I think you will be surprised.

As I said in my first post, everything in audio is a compromise. On the speaker design side, more complex XO's are just that, a compromise you use to control drivers and their responses. You play the cards you are dealt. It just doesn't fit into my Utopian view of what audio should be. Then again, I seem to 'buck' the norm more often than not.

Every speaker I build now is bi or tri amped at some point during the design process, I have a pair of those cool Vidsonik crossover boxes that AudioExpress is selling.  But since getting them I've been using drivers that need a big of response shaping, like most that give any real detail do.  I haven't tried to do something with flat as possible woofers and tweeters lately, but I have a couple perfect projects in mind, and the drivers here.  It might be a fun experiment to try and report back here.  I'm betting on my crossover though 

[Scott F.]

big snip.......  I haven't tried to do something with flat as possible woofers and tweeters lately, but I have a couple perfect projects in mind, and the drivers here.  It might be a fun experiment to try and report back here.  I'm betting on my crossover though 

Cool. Lets plan on it later this summer. You might be right about your XO but keep an open mind. I think you'll have a very pleasant surprise when you hear how a good active sounds. Maybe by then I'll have my AR EC-3 tubed XO back. That thing is extremely cool but not very flexible. It has a fixed 12db slope though there are plenty of XO points.

[Vapor Audio]

big snip.......  I haven't tried to do something with flat as possible woofers and tweeters lately, but I have a couple perfect projects in mind, and the drivers here.  It might be a fun experiment to try and report back here.  I'm betting on my crossover though 

Cool. Lets plan on it later this summer. You might be right about your XO but keep an open mind. I think you'll have a very pleasant surprise when you hear how a good active sounds. Maybe by then I'll have my AR EC-3 tubed XO back. That thing is extremely cool but not very flexible. It has a fixed 12db slope though there are plenty of XO points.

Open mind?  I've been on audiocircle way to long for that 

[*Scotty*]

TurboFC3S, Do you feel there is any sonic penalty to be paid when you have use response shaping elements and notch filters in a design in order to hit the desired response characteristics
compared to a simpler xo design with lower parts count.
Scotty

[Vapor Audio]

TurboFC3S, Do you feel there is any sonic penalty to be paid when you have use response shaping elements and notch filters in a design in order to hit the desired response characteristics
compared to a simpler xo design with lower parts count.
Scotty

Well, there are a lot of permutations to that question.  Assuming a circuit is applied for the right reasons, and all the important factors are weighed when doing so, the benefits are obvious - and any penalty is debatable.  Any time I use something like a notch, I listen with and without.  If the benefit isn't clear, I usually don't keep it.  I almost always do a full horizontal axis measurement sweep with/without, also measure phase, step, burst decay, etc ... I have to see or hear a clear reason to keep any extra components in the signal path.  Personally I don't design for flat response as many do, in fact on axis gated FR is not given very high priority by me at all.  Many internet designers use flat as their final goal, it looks great posting a ruler flat measurement on some bulletin board, and they'll add layers of unnecessary complexity to get there.  And then they don't even listen until the end, after it's all soldered up and say, "Hey, it sounds pretty good".  But they have no idea some of the revisions they threw away on the path to the flat FR would have sounded better. 

Point is, it's my job as a designer to make SURE there's no penalty to added circuits. 

[TONEPUB]

While the finest system I heard at RMAF 2007 and CES 2008 (the Sonicweld Pulserod) used a hotrodded DSP crossover, achieving that system's level of transparency is not trivial and so I am not sure that it would be cost-effective in most situations.

If one only knew the dedication to his craft, time spent perfecting said craft, and in-depth digital and DSP design knowledge that Josh Heiner (Sonicweld) has, they would quickly realize that utilizing an active crossover with DSP it is not a trivial undertaking or cost-effective solution whatsoever.  However, to my ears, his speakers produce the finest reproduction of recorded music I've ever heard.

He is slammed building product, doing high end medical and military design projects, so I doubt seriously he would ever find the time to post here, but visit our room at RMAF and feel free to ask him how much work and how many years he has invested to be at the point he is, you'll get an earful!   

Disclaimer--Josh is my closest friend and I have watched his growth as a designer over the last 15+ years, so call me biased if you'd like. 

Peace,

Lee     

We lived with the Sonicweld system for quite a while and I was very impressed at what Josh has accomplished.
While not my absolute favorite speaker on Earth (one of the top 5 though), certainly the best active speaker I've heard by a considerable margin, and definitely the best Sub/Satellite integration I've ever heard...

[andyr]

As Roger suggested, the really interesting "tweener" (but largely ignored) is the passive line level, maybe with buffers.  There have been some versions over the years, but usually implemented with lousy sounding op amps and cheezy power supplies.

Sorry, Tom, not sure whether you had a "senior moment" or simply too much to drink before you posted but a PLLC (passive line level crossover) consists of resistors and capacitors - ie. it is passive!    An active line level crossover (a normal "active XO") uses opamps or transistors and, hence, needs a PS.

[andyr]

True you don't need a Zobel with active xovers, ... 

I'm afraid I disagree.  The Zobel is there to level out the driver's impedance - and it has to be "attached" to the driver, to work.  In a passive XO, the Zobel enables the filter to work as theoretically calculated (based on a constant driver impedance); in an active setup, it stops the amp delivering varying output levels into different driver impedances.

The main advantage I see to an active network and multiple amps is being able to eliminate back-emf, but that's not exactly a major concern.

I wouldn't agree that eliminating back-EMF is necessarily a minor concern but the chief advantages of active filters, IMO are:
* they allow the amps to directly control the driver ... instead of putting hundreds of feet of wire (which make up the LP inductors) in between the amp & the driver.
* they allow multiple lower-power amps to be used, instead of a single high-power amp.  Typically - as I understand it - if two amps have a similar topology, the one with multiple output devices will not sound as refined as the one with only one output device per channel (as it is difficult to keep the multiple output devices working in synch).

Regards,

Andy

PS:  I converted my Maggie IIIas into 3-way active about 10 years ago.    Luckily, Magnepan drivers are almost entirely resistive.

[andyr]

This one seems to be a double post?