[Occam]

Hey Andy,

A few minor points -

Active multi-amping doen't necessarily obviate the need for all passive components betwixt and between the amp and drivers. Often the rising impedance of a driver would still mandate the a Zoebel, or other components, for compensation.

While I'd generally agree that all things being equal, a smaller amp often sounds better than a larger one, there are situations where all things are definitely not equal. Specifically, balanced bridge topologies, depending on the ps can offer a quadrupling of power output and improved sonics; the McCormick DNA500 vs their smaller bretheren for example. Similarly, I'd imagine the benefits of vertically bi-amping a Belles Signature 150/350 single ended vs single amping with the same in balanced bridged configuration would not be so clear cut.

The nature of the speaker level crossover would also impact the optimal choice. I'm unsure as to the benefit of multi-amping very simple crossovers such as those on Epos and Green Mountain Audio speaks. These typically have 6db (or less, often depending on the driver's characteristics) electical crossovers.

In situations with complex crossovers, specifically those that provide more than band splitting duties, the choice can certainly swing towards active multi-amping. SL's Orion comes to mind. The fact that the impedances dealt with in active line level crossovers are far higher than those in speaker level crossovers can swing the choice towards active. Depending on ones own specific cost benefit decisions, a choice of bespoke capacitors (teflon, silver foil in oil) could make the decision to actively multi-amp actually more economical. A .033uf teflon/foil caps is far less expensive than a 20uf one. And as one use more complex slopes / frequency manipulations, with greater numbers of components, the savings offered by using smaller values dictated by higher impedances of line level, could pay for the addional amplifiers.

And obviously, multi-amping provides us the opportunity to voice our system, such as using ss for the bass and hollow state for the highs. And for the truly adventurous and masochistic, it provides an opportunity to skate on the bleeding edge with regards to compensation and feedback. We could use an amp specifically for bass that would self destruct if asked to power our treble, and versa visa  
So many ways to blow up our systems, and so little time....

[Davey]

Andy,

You'll recall that we had a discussion of passive-biamping on the Planar forum awhile back.  After that I thought you had it pretty well straight in your head.  It appears you still do.  

The question of the original post seems simple upon first reading, (and the simple answer would be "no") but it can get a bit more complicated depending upon a persons understanding of how a power amplifier supplies current to the load.
The concept is fairly simple, but it boils down to this.....you can have voltage without current, but you can't have current without voltage.  In a passive biamping system the load resistance/impedance goes very high in the out-of-band area...(high frequencies for the low-pass section and low frequencies for the high-pass section.)  In these areas the amplifier has the voltage "available" (since it's still be fed with a full-range signal) but it doesn't have to "supply" any current to the load at those frequencies.  Does this mean that each amplifer is "wasting power?"  No, not in my mind.  Could each amplifier be used in a more efficient way?  Yes, with a line level crossover.    I know that doesn't make logical sense for some folks, but I can't explain it any better.    A discussion of this situation can get into semantics pretty easily...and of course, what is the definition of "waste."

Heck, my view of "passive bi-amping" is that it's not even really bi-amping.  The traditional concept (which dates back many many years) of bi-amping was always a direct connection between speaker and amplifer and the crossover shifted to line-level.  The definition has been skewed in recent years by "audiophiles" who have no understanding of the technical concepts involved.  For them, if they have TWO amplifiers they're bi-amping.  

Heck, the definition of "audiophile" has been completely bastardized in the past few decades.....but that's another topic.  

Cheers,

Davey.

[andyr]

... Heck, my view of "passive bi-amping" is that it's not even really bi-amping.  The traditional concept (which dates back many many years) of bi-amping was always a direct connection between speaker and amplifer and the crossover shifted to line-level. The definition has been skewed in recent years by "audiophiles" who have no understanding of the technical concepts involved. For them, if they have TWO amplifiers they're bi-amping. ...

Hi Davey,

It's been a long time since I passively bi-amped my IIIAs, so I've forgotten precisely what the difference was when I moved from one amp to two - yeah, maybe it didn't make much difference.    But I'd paid so much money for the second Naim 250 that it had to be better!!  

However, I was amazed that my doctor mates noticed this huge difference in sound quality when they used two Brystons together with (the passive crossover in) their Dalis.

Like night & day, they said!     And there was a big difference between using one Bryston per Dali, and running the highs on one and the lows on the other (but I can't remember which way they said was better!).  One way didn't make much difference - the other did!!

So that "proves' to me that passive bi-amping can be better than just using one!  

... Heck, the definition of "audiophile" has been completely bastardized in the past few decades.....but that's another topic...

I thought an "audiophile" was someone whose ideal evening's listening was playing the same test disc through different components in his system ("his" because women don't ever seem to want to do this!    ) ... whereas a music lover would rather spend the evening listening to several discs right through!!  

Regards from God's Own,

Andy

[andyr]

... A few minor points -

Active multi-amping doen't necessarily obviate the need for all passive components betwixt and between the amp and drivers. Often the rising impedance of a driver would still mandate the a Zoebel, or other components, for compensation. ...

Hi Paul,

I agree that there is often good reason to put a compensation network on the drivers when using an active crossover ... but I thought the Zoebel was simply in parallel with the driver - so, with no series component involved, the amp still has direct control of the driver?  Only when you introduce - in particular - an inductor, do you degrade from "direct control" to "fly-by-wire" (to use an old flying analogy).

"ss for the bass and hollow state for the treble"  haha - nice dig!!     But I must tell you, listening to these monstrous Dalis the other evening through firstly the Bryston and then through Hugh's 40w SET ... the "hollow state" gave a stupendous soundstage and engagement (compared to the boring Bryston).  

And a friend who came round to listen to my Maggie IIIAs yesterday commented how he'd auditioned some 3.6s at a dealer and they sounded boring with the amp which the dealer was using but then (after my friend commented on this) he brought out a 30w SET and "magik" appeared!!  

Regards,

Andy

[Occam]

Andy,

Passive components in shunt positions are still in the signal path, even if only to the extent that their effect is subtractive. One can (although I doubt that I could) make an orthogonal transform [EDIT: Ooops, that should be isomorphic transform] from many parallel to series crossovers of arbitrary complexity, for example. But looking at the extent that a shunt Zoebel would realistically effect sound might yeild a minor effect under normal circumstances. But in situations at the extreame, like with an amp with bleeding edge compensation/ feedback schemes, or anywhere where alphacore cables (extremely capacitive) are used, the effect could be quite material.

My use of the term 'hollow state' wasn't meant by any means to be a dig. If I had my druthers, and an extremely efficient bass driver, and lots of money, my preference would be to drive it with a Atma-Sphere OTL amp. It has outstanding bass under the right circumstances. My use of 'hollow state' was simply a reference of principal of transmission of electrons through a vacuum, hence hollow, as opposed to solid state, which operates via the principal of keeping the smoke in. Everyone knows that if the smoke escapes, it no longer functions.  

[avahifi]

To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system.

Does your electronic crossover provide for the same pole points and slopes that the speaker designer wanted for the drivers used?  Are you sure of what specifications are really needed? Can you really set all the controls on your electronic crossover to get these specifications.  Remember what it is doing is not what the knob is pointed at, that is only an approximation.  You need a dual trace scope and signal generator to get the slopes set just right and get the gain equal for both channels.  Do you have that equipment available?

Does your electronic crossover provide for whatever frequency response "tweeks" that your speaker designer built into the design to make up for speaker characteristics that he knew about but maybe you didn't?

What about the active circuits in your electronic crossover?  It is really another preamplifier in series with your system you know.  You love tube sound but have an all IC electronic crossover?  Huh?  Lots of ICs likely, lots of inexpensive ICs that you would never tolerate in your main audio preamp I suspect.  But now you have them.

My suggestion for the best way to eliminate all these variables is to use a great loudspeaker system that can be driven full range with a great amplifier and use a great full range amplifier.

Frank Van Alstine

[Tweaker]

To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system...

 These considerations have prevented me from going that route. Not to mention the fact that additional interconnects would have to be introduced into the chain.

[Davey]

Frank,

Obviously, (or maybe not to some folks) an active crossover would have to be designed with the same slopes that a passive design implements in order for the "design" to be the "same."  (If that was indeed the design objective.)  This type of "duplication" is rather easy to accomplish these days with many speaker design programs that allow optimizing to a "target."  Whether or not that is the best approach is up for debate, but the tools to do it are here.  I've done this on many occasions.

However, many systems nowadays are designed from the ground up with an active crossover....ala Linkwitz Orion.  Other folks (like Andy) are experimenting with active crossovers as replacements, but they are aware of the fact that the slopes may or may not be the same.  They factor that into their evaluations and decisions.

Active circuits.....well designed....are essentially transparent and contribute much less non-linearity, distortion, etc, etc, than a high-level speaker crossover would.

Tweaker,

If you're worried about additional interconnects in the chain then an active crossover setup wouldn't be for you.  

Cheers,

Davey.

[andyr]

Andy,

Passive components in shunt positions are still in the signal path, even if only to the extent that their effect is subtractive. One can (although I doubt that I could) make an orthogonal transform from many parallel to series crossovers of arbitrary complexity, for example. ...

Hi Paul,

Of course the Zoebel network "affects the sound" (as I understand it, being an electronics weeny, it simply makes the frequency response of an inductive driver more "flat") but I don't believe it changes the "iron-fisted control" which an amp has over the driver in an active crossover setup (compared to a passive crossover).

And - even though I'm an Aussie - I got your "electron" reference wrt "hollow state"!     That's a good term - I think more apposite than "glass amp"!!  

Regards,

Andy

[andyr]

To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system.

Does your electronic crossover provide for the same pole points and slopes that the speaker designer wanted for the drivers used? Are you sure of what specifications are really needed? Can you really set all the controls on your electronic crossover to get these specifications. Remember what it is doing is not what the knob is pointed at, that is only an approximation. You need a dual trace scope and signal generator to get the slopes set just right and get the gain equal for both channels. Do you have that equipment available?

Does your electronic crossover provide for whatever frequency response "tweeks" that your speaker designer built into the design to make up for speaker characteristics that he knew about but maybe you didn't?

What about the active circuits in your electronic crossover? It is really another preamplifier in series with your system you know. You love tube sound but have an all IC electronic crossover? Huh? Lots of ICs likely, lots of inexpensive ICs that you would never tolerate in your main audio preamp I suspect. But now you have them.

My suggestion for the best way to eliminate all these variables is to use a great loudspeaker system that can be driven full range with a great amplifier and use a great full range amplifier.
 ...

Hi Frank,

It's a pity that someone of your reputation is introducing furphys into this discussion.

You said: "To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system."

Well ... yes!!  Exactly!!  And you are removing a passive crossover!!  

Neither crossover is "100% transparent"!!  The passive crossover introduces phase shifts and sonic degradation ... and so does the active crossover.

Then you said: "Does your electronic crossover provide for the same pole points and slopes that the speaker designer wanted for the drivers used? Are you sure of what specifications are really needed? Can you really set all the controls on your electronic crossover to get these specifications."

Good points ... in my case, yes, my Maggie IIIAs sound better with a 3-way active crossover using exactly the same slopes as the stock passive crossover.  (I tried 24dB slopes too.)

How did I do this? ... I used lspCAD to model the slopes and -3dB & -6dB points of the stock passive crossover, and then I tweaked Rod Elliott's active crossover circuit to deliver the same slopes (18/12dB for bass/mid & 12/12dB for mid/ribbon).  Except I slightly spread apart the -3dB points to deliver a much flatter overall (electrical) frequency response ... as I figured Magnepan didn't have that kind of computer technology available in the late 80s!!

And then I used a sig-gen and a CRO to set each band at the correct gain.

The nett result - on music with fast scales, like Beethoven's "Appassionata" at 45rpm - seamless driver integration ... no-one can pick when each driver passes over to the next.

And then your next point: "What about the active circuits in your electronic crossover? It is really another preamplifier in series with your system you know. You love tube sound but have an all IC electronic crossover? Huh? Lots of ICs likely, lots of inexpensive ICs that you would never tolerate in your main audio preamp I suspect. But now you have them."

Well, yes ... an active crossover is just another gain stage (even if it's operating at unity gain!    ).  But it gains by not having the nasty inductors in the signal path which a passive crossover has - nor huge capacitors  (you can buy high quality caps in the 100nF or less range for your active crossover but you can't buy them in the 150uF range!!)

And yes, ICs have their own sound - so you have to select good ones - and they ain't cheap!  But then again, if you choose some expensive good quality opamps, they will most probably sound much better than the commercial dross used in most mixing consoles ... of course, different matter if you are listening to Tacet, all-tube mastered LPs ... but there ain't many of these!  

Finally, you say: "My suggestion for the best way to eliminate all these variables is to use a great loudspeaker system that can be driven full range with a great amplifier and use a great full range amplifier."

Really, Frank, I'm surprised you didn't say "use a great single-driver full range speaker system" ... as everyone knows any crossover introduces colouration!

And wrt using one "great amplifier" vs. using, say, 3 smaller amps in a 3-way active system ... I suggest that most of the time, amps which are built to deliver 800wpc (into 4 ohms) do not sound as refined as amps which are built to deliver, say, 200wpc, 110wpc & 50wpc into 4 ohms (which is what I am using).  The multiple pairs of output transistors needed to deliver huge amounts of power have their sonic downside, compared to smaller-powered amps.

Regards,

Andy

[andyr]

To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system...

 These considerations have prevented me from going that route. Not to mention the fact that additional interconnects would have to be introduced into the chain.

Hi Tweaker,

I presume you have an integrated amp then?  (So you can avoid using an IC!)

regards,

Andy

[avahifi]

I make a few factual observations, and get answers with a lot of opinions.

That's normal around here. Oh well.

Anyway, lets leave the cables out of the mix shall we, along with our good sounding parts. Or I could mention that if you use an electronic crossover and multple amplifiers, now you have several more evil sounding power cords to deal with.

Anyone have any actual data that well designed passive speaker crossovers sound bad?  And yes I have heard "one way" systems, one a very special experimental design that was quite impressive.  However the range limitation did not convince me that eliminating the crossover was a good trade off.

How do you select good sounding ICs?  Do you know their peak current capacity as compared with the amount of current required to drive the capacitive load of the filters designed around them?  Our observations are that no linear op-amp by itself can manage that kind of load. Ever consider the thermal intermodulation distortion caused by the heat transfer from the current amplifer part of the IC die back to the voltage amplifier portion? Most don't even know this problem exists.

There is no good reason for a well designed big amplifier to sound any worse that a small amplifier. Our OmegaStar amps, for example, are  identical whether we build the 90 watt version or the 400 watt per channel version into a hafler p500 shell - - which is very impressive by the way.

You might have the test equipment to actually match an active crossover to you speaker's needs, but the average person will not.

Hey, I should be all for electronic crossovers and the use of lots and lots of power amplifiers, it would help me sell more power amplifiers.  However, I am first for helping my clients get convincingly musical audio with reasonable equipment and realistic setup and use requirements.  To me the KISS theory is the way to go (Keep It Simple, Stupid).

Frank Van  Alstine

[konut]

I'm glad to see some authoritative views on the issues discussed here. While I won't pretend to understand all the details of the different issues, it does reinforce some of the choices I've made. Any crossover, be it passive, active, power level, or line level has problems associated with it that can be mitigated, to a certain degree, by careful implementation.
      I use a 'single driver' speaker, the Omega Aperiodic 8. The advantage of this is that it removes the crossover from the midrange where the ear is most sensitive. Of course its not full range. I used an old Behringer active xover to cross at 90hz to my sub but, for the reasons Frank pointed out, it mangled the signal to a degree. I have since upgraded to a Marchand XM46 passive line level high pass filter at 70hz. This has restored the transparency I had before using the sub and active xover. I'm also using a passive attenuator with no preamp. The variable  filter in the sub handles the low pass duties. Hows that for keeping it simple Frank?  

[Occam]

I make a few factual observations, and get answers with a lot of opinions.
That's normal around here. Oh well.
........
How do you select good sounding ICs? Do you know their peak current capacity as compared with the amount of current required to drive the capacitive load of the filters designed around them? Our observations are that no linear op-amp by itself can manage that kind of load. Ever consider the thermal intermodulation distortion caused by the heat transfer from the current amplifer part of the IC die back to the voltage amplifier portion? Most don't even know this problem exists. ...

Frank,

I accord you a certain amount of respect by virtue of the quality of your products and your longevity in the business. But please understand, I would not defer to you the same way I would to a John Curl ( or a Fred Deikman, for that matter). What you assert are facts are simply what I would call viable solutions. There is inevitably, more than one way to skin a cat. What you assert as 'facts' without what is necessary and sufficient to prove such, is simply bluster. Well considered bluster no doubt, but nevertheless, bluster.

Certainly, the willy nilly substitution of IC is often unwise. One should make the decision as to what ICs might be suitable for a given application the same way one does so in choosing discrete circuitry. What is the source impedance, load impedance.... What voltage swing does one need, what noise is acceptable, and what output offset? In other words, you read and understand the frigg'n datasheet. If one is concerned over non constant heat is a output stage biased AB, you bias it into class A for the given swing and load....  And certainly ps decoupling is critical when considering the use of high gain/bandwidth opamps. After one satisfies the technical needs, then one can explore subjective preferences. Frankly, I find your assertion that IC are not capable of driving properly designed filters (or incorporated into the feedback loop) ludicrious.

Feel free to be as patronizing as you wish in your own Circle. But please, don't get your knickers in a twist when all to not kowtow before you. Would folks probably be better off to buy a proven full range speaker and an excellent amplifier to power it?.... probably yes. But if someone wishes to implement a speaker designed (or design their own) to use multi-amping, realize that there are no hard and fast rules when it comes to subjective preferences.

[Dan Banquer]

All you have to do is turn on just about linear amplifier and it will waste more power at idle than what it puts out to the speaker. Typical stereo amplifiers will consume anywhere from 50 watts to 120 watts at idle. Given that most of us average one watt to the speakers I think that's lopsided enough.
Hope this helps;
                d.b.

[warnerwh]

Theoretically how much difference is there between using an active crossover and a passive crossover?  I'd like to know measured differences so we could have some idea of what may be audible.

Of course this wouldn't include as Frank brought up about what a speaker maker may put into the passive crossover to correct for various anomolies that wouldn't be taken care by most people when active biamping.  These parts would more than likely make a significant and audible difference presumably for the better, either that or the speaker designer is an idiot.

Dan would you knock it off with the good points   No doubt running two amps like I am the amount of waste power is significant as both amps put out some heat even at low volume.  In Oregon it's good thing to have heat most of the year though.

[_scotty_]

warnerwh asked,

Theoretically how much difference is there between using an active crossover and a passive crossover? I'd like to know measured differences so we could have some idea of what may be audible.

Just speaking for myself, I don't understand what" measured differences"
you are referring to in your question.  
Scotty

[Tweaker]

I presume you have an integrated amp then? (So you can avoid using an IC!)

Oh, Lord.
No, I mentioned it because there are many of us who try hard to minimize the number of componants between the source and speakers as every one, we believe, including interconnects has the potential to further color the sound. Then there is the added cost, as, because not every one has Radio Shack cables, the additional expense to buy a couple of additional pairs could be very significant. Perhaps more so than the resulting improvement in sound, assuming there is any.

[warnerwh]

Scotty, what I am referring to is anything that can be measured that can affect the sound. If there's nothing to measure that can affect the sound I guess it's a toss up.  What I don't understand is why so many people say it's not worth biamping unless you go active.

Maybe I should have asked how can one type of biamping sound better than the other as far as sound quality is concerned.  I would like a detailed reason for this.   Thanks

Maybe measuring the signal at each speaker driver would be different with one type of crossover over the other.  I don't know or understand and that's why I'm asking. I'm just curious. I have no intention of going active especially if the only benefit is having an extra 6db or whatever it is in output.  My amps are overkill already.

[Occam]

To muddy the waters even further, when you go for active bi-amping, now you are introducing an electronic crossover into the system...
Huh? Lots of ICs likely, lots of inexpensive ICs that you would never tolerate in your main audio preamp I suspect. But now you have them. .....

These considerations have prevented me from going that route. Not to mention the fact that additional interconnects would have to be introduced into the chain.

Itaicized script added from Frank's post, by me.

Huh? I sadly have to tell you that given your described system -
http://www.audiocircle.com/index.php?action=systems;system=575
that the Behringer DEQ2496 you state that you've inserted in your system adds 14 opamps, 7/channel, via JRC4580 'jellybean' ICs into your circuit path. I'm certainly not disputing your choice, and obviously you feel the benefits offered, including the extra interconnects, results in a net gain in audio quality..... just like many who use IC based [but certainly better that those POS JRC opamps] line level crossovers feel that those compromises are similarly worth it.

Perhaps Frank, with his vast storehouse of definitive facts, could provide you some guidance in upgrading those crap IC opamps. He could also help you address that ever present bugaboo of thermal intermodulation distortion, that concerns him so deeply.