What are the pro's and con's of horizontal biamping with say a sweeter less powerful tube amp on the mids/treble and a beefy SS taking care of bass duties?
Anyone running such a setup? What are the watch-outs (besides gain mismatch obviously)?
Phase missmatch, one amp will be on one phase and the other not quite, it will just sound bad... If you have a way of adjusting phase, either a button on your xover or time aligning the drivers it can be fixed.
If you are crossing low enough you might get away with it too, say under 100 hz.
You will get a different sound which you can like for the first couple of days, when you finally get back to just one amp and everything falls into place again you will see what I mean!
I am running the top half of my speaker with a pair of Cary CAD-211's, and a solid state amp for bass duties. I have also taken the additional step of bypassing the internal crossover of the speaker and using an active external crossover - in other words, the amp is directly connected to the driver with no crossover in between. The improvement in control and transparency is simply massive.
Before I did this, I did run a simple horizontal bi-amp setup with an additional passive preamp for the bass section to provide level matching. It did yield improvements over the single amp setup (more dynamic headroom, better control of the bass), but the real gains came from going fully active.
I did this for years with my Magnepans, running a big Counterpoint amp on the bass panels and tube ARC amp on the mid-tweeters. As Saturntube said, gain and phase must be carefully matched at the crossover point(s). Modern DSP-based crossover make that easy, though you must still measure to optimize the system, or you can build an analog phase-adjust (all-pass) circuit. As Keith_W said, the biggest advantage in bi-amping is replacing the crossover networks, but that also takes the most work, the most technical savvy, the most measurement equipment, and the conviction that you can do a better job than the speaker designer.
As Keith_W said, the biggest advantage in bi-amping is replacing the crossover networks, but that also takes the most work, the most technical savvy, the most measurement equipment, and the conviction that you can do a better job than the speaker designer.
Well I certainly didn't think I was smarter than the speaker designer when I replaced my crossover network! As a starting point I simply replicated the speaker crossover settings in my new active crossover. From there I took quite a few measurements, got the help of a local speaker designer, and came up with refinements of our own. Some of them didn't work too well, e.g. going with a very steep crossover slope was a backwards step, but overall it was a step forward.
Bypassing the internal passive crossover and utilizing an external active crossover (Pass Labs XVR1 three-way) and running separate amplifiers for woofers, midranges and tweeters made a tremendous improvement (at least in what constitutes a "tremendous" improvement in high-end tweaking). I also started with the same crossover points and slopes as in the original speaker specs, but I tweaked it slightly for the better.
Having the power amps directly connected to the speaker drivers instead of going through a passive crossover makes a big difference--MUCH bigger difference than any speaker cable upgrade.
Good points above. I also use a Pass Labs XVR-1 3 way set up using 3 different amplifiers. The typical veiling of a passive crossover can be remedied by going active and it makes for a noticeable improvement in transparency and dynamics. That's a very big plus. It may be all that matters depending on your system set-up. Critical listening over time may reveal tradeoffs - a less integrated soundstage and somewhat poorer imaging. This is unlikely to be a problem if you're adding a subwoofer with a low crossover frequency. It's more problematic at crossover frequencies above 100HZ where sound becomes directional. It's an even worse problem when going from bi to tri-amping. Note, even with the Pass Labs' exceptional flexibility, I can not exactly duplicate my Magnepan 3.6's crossover Fqs, slopes, and phase, its electrical response. Fortunately it's the in-room acoustic response that matters and that's usually an entirely different animal. As mentioned, you will find exact crossover channel gain matching is very critical as imbalances of less than a dB, especially in the mid-range, will really upset imaging. Despite all this, I have been able to get very good results with a lot of crossover fiddling in conjunction with extensive in-room measurements so I wouldn't let any of this dissuade you from experimenting. Adding a subwoofer is easy and effective. Some speakers have less complex crossovers. Both the manufacturers or your speakers and crossovers can give you good advice on how to get the best results. You may even be able to set all this by ear with a minimum of effort and be quite happy with the results!
Well I certainly didn't think I was smarter than the speaker designer when I replaced my crossover network! As a starting point I simply replicated the speaker crossover settings in my new active crossover. From there I took quite a few measurements, got the help of a local speaker designer, and came up with refinements of our own. Some of them didn't work too well, e.g. going with a very steep crossover slope was a backwards step, but overall it was a step forward.
Yeah, that was a little too strong... I did the same with my Maggies, started with the factory crossover and slope, then tweaked from there. Same for other speakers I piddled with. I do think you can get yourself into trouble if you are not cognizant of what the drivers can do and have measuring equipment on hand. I am sure we all have tales for folk (not me, the other guy, of course) who tweaked by sound and discovered while the SPL went up crossing the woofer over an octave or three higher, so did distortion, and dropping the midrange driver's crossover might provide "cleaner" bass but tended to shorten the lifetime of the little midrange driver somewhat...
Getting the phase right was a pain and for me usually the hardest part of the job. It wasn't hard to get close as the internal crossovers were usually pretty low-order so I could duplicate them at line level, but tweaking to optimize the impulse/pulse response was a challenge at times.
What are the pro's and con's of horizontal bi-amping with say a sweeter less powerful tube amp on the mids/treble and a beefy SS taking care of bass duties?
Are you talking about true bi-amping with an active crossover or passive bi-amp?? The biggest concern going active is correctly matching the voltage drives of the passive crossover with an active. With almost all analog crossovers the stock choices are limiting. Many systems don't use textbook crossover slopes and many also have some kind of filtering on all or some of the drivers.
What you need to do is evaluate the passive network to see exactly what it's doing. Once you understand this you can go from there. If it has simple generic slopes and no notch filters you can take a look at what's available in the analog domain that might be a good match. There are not that many out there. Some are card based, some are generic slopes with a simply dial frequency select others are set-up using switches and jumpers. You can also go to Marshland for custom cards if needed.
If you can't find a good analog fit you can go digital where you have almost infinite possibilities and you can match the most demanding voltage drives, adjust delay and so on.
It used to be you could find a number of analog active crossovers offering independent control of gain, slope, and crossover, and some with phase control. Their numbers seem to have dwindled over the years. A DSP-based solution certainly offers much greater flexibility, especially in phasing, and stability.
I was actually thinking about passive bi-amping. To get specific, a DartZeel nhb108 for mid/treble and burmester 911 on bass on the tad cr1's. The dart alone doesn't quite have the iron first control on the woofer the 911 has with its damping factor. I am not ready to mess around and try to bypass the internal crossover so it will be passive for the time being. Fully understand the technical merits to active biamping though.
Hmmm... May need to define pasive bi-amping... Please consider all that follows In My Opinion.
Active bi-amping = line-level crossover, each power amplifier thus sees only a limited range of frequencies. The line-level crossover could actually be a passive network. The main benefit is to offload the bass from the treble amp and vice-versa, which in turns allows one to choose the appropriate amp (power, type, etc.) for each frequency range. Examples might be a high-power amp on the woofer and less-powerful amp on the mid/treble. IME amps in the audiophile world are often chosen for their "character" in this configuration; I chose a tube mid/treble amp and hybrid (tube input, FET output) amp for the bass driving my Magnepans. Pro systems I have set up and used tend to use similar amps, often the same manufacturer, but at different power levels. At least partly to keep from blowing the club's breaker every five minutes...
Passive bi-amping = as implemented in most AVRs/pre-pros/SSPs, simply applies the same signal to each power amp and keeps the speaker crossovers (requires speakers that are bi-wire/bi-amp capable). This offloads some of the power requirements but not voltage swing and so is of questionable (and thus arguable, and much argument has been done and will no-doubt continue) benefit.
Active bi-amping provides the option of removing the speaker's internal crossovers to optimize the performance; that is the next step.
A simple 1st or 2nd order passive crossover in front of the power amps might be a good place to start if you want to try bi-amping.
I'm doing it without an external crossover. Seamless because input sensitivities of the different pair of amps are almost identical ( difference less than .02).
Hmmm... May need to define pasive bi-amping... Please consider all that follows In My Opinion.
Active bi-amping = line-level crossover, each power amplifier thus sees only a limited range of frequencies. The line-level crossover could actually be a passive network. The main benefit is to offload the bass from the treble amp and vice-versa, which in turns allows one to choose the appropriate amp (power, type, etc.) for each frequency range. Examples might be a high-power amp on the woofer and less-powerful amp on the mid/treble. IME amps in the audiophile world are often chosen for their "character" in this configuration; I chose a tube mid/treble amp and hybrid (tube input, FET output) amp for the bass driving my Magnepans. Pro systems I have set up and used tend to use similar amps, often the same manufacturer, but at different power levels. At least partly to keep from blowing the club's breaker every five minutes...
Passive bi-amping = as implemented in most AVRs/pre-pros/SSPs, simply applies the same signal to each power amp and keeps the speaker crossovers (requires speakers that are bi-wire/bi-amp capable). This offloads some of the power requirements but not voltage swing and so is of questionable (and thus arguable, and much argument has been done and will no-doubt continue) benefit.
Active bi-amping provides the option of removing the speaker's internal crossovers to optimize the performance; that is the next step.
A simple 1st or 2nd order passive crossover in front of the power amps might be a good place to start if you want to try bi-amping.
A good summary, but I'm not sure I agree with the main benefit to active. Getting around the speakers internal crossovers and, therefore, having the amps directly in control of the drivers is pretty critical, IMO.
Hmmm... May need to define pasive bi-amping... Please consider all that follows In My Opinion.
Active bi-amping = line-level crossover, each power amplifier thus sees only a limited range of frequencies. The line-level crossover could actually be a passive network. The main benefit is to offload the bass from the treble amp and vice-versa, which in turns allows one to choose the appropriate amp (power, type, etc.) for each frequency range. Examples might be a high-power amp on the woofer and less-powerful amp on the mid/treble. IME amps in the audiophile world are often chosen for their "character" in this configuration; I chose a tube mid/treble amp and hybrid (tube input, FET output) amp for the bass driving my Magnepans. Pro systems I have set up and used tend to use similar amps, often the same manufacturer, but at different power levels. At least partly to keep from blowing the club's breaker every five minutes...
Passive bi-amping = as implemented in most AVRs/pre-pros/SSPs, simply applies the same signal to each power amp and keeps the speaker crossovers (requires speakers that are bi-wire/bi-amp capable). This offloads some of the power requirements but not voltage swing and so is of questionable (and thus arguable, and much argument has been done and will no-doubt continue) benefit.
Active bi-amping provides the option of removing the speaker's internal crossovers to optimize the performance; that is the next step.
A simple 1st or 2nd order passive crossover in front of the power amps might be a good place to start if you want to try bi-amping.
NO Benefit except to the electrical power company and to your mind if it makes you feel better.
If the amp is getting a wider signal range than the passive crossover in the speaker cabinet sends to the driver, the power is being wasted in the passive crossover and the amplifier is being overworked and having more intermodulation distortion than it would if the active crossover was used between the preamp and power amp and the amp directly powered the speaker driver giving it only the frequency range it would reproduce.
Having the amplifier directly controlling the speaker driver without an intervening passive network makes a tremendous improvement in sound. It is "pretty critical" as you say and not for the timid, but most speaker crossover networks are easily accessible by taking off a rear panel or a woofer to make the disconnections. Just snip all the leads going to the drivers and solder on a new piece of wire to connect to your power amps.
You can always reconnect the drivers to your passive crossover if you don't like it, which is highly unlikely. Don't worry, you'll find something to do with those extra power amps you bought!
NO Benefit except to the electrical power company and to your mind if it makes you feel better.
If the amp is getting a wider signal range than the passive crossover in the speaker cabinet sends to the driver, the power is being wasted in the passive crossover and the amplifier is being overworked and having more intermodulation distortion than it would if the active crossover was used between the preamp and power amp and the amp directly powered the speaker driver giving it only the frequency range it would reproduce.
Having the amplifier directly controlling the speaker driver without an intervening passive network makes a tremendous improvement in sound. It is "pretty critical" as you say and not for the timid, but most speaker crossover networks are easily accessible by taking off a rear panel or a woofer to make the disconnections. Just snip all the leads going to the drivers and solder on a new piece of wire to connect to your power amps.
You can always reconnect the drivers to your passive crossover if you don't like it, which is highly unlikely. Don't worry, you'll find something to do with those extra power amps you bought!
Most speakers have complex crossovers and their transfer functions can not be emulated using the typical active crossovers, such as the Berhinger DCX2496 or Marchand's, that only allows the selection of a few basic filter shapes, slopes and phases. Should we expect that an approximate transfer function is better than the original, carefully studied and optimized by the designer?
I was actually thinking about passive bi-amping. To get specific, a DartZeel nhb108 for mid/treble and burmester 911 on bass on the tad cr1's. The dart alone doesn't quite have the iron first control on the woofer the 911 has with its damping factor. I am not ready to mess around and try to bypass the internal crossover so it will be passive for the time being. Fully understand the technical merits to active biamping though.
What is the crossover frequency you are aiming to? I have tried this approach in the past and the result was very poor - the blending of the amplifiers was very imperfect. Said otherwise - there was no coherency in the system, some parts of the spectrum were clearly heard as too "electronic" and "mechanical".
My preference for by-amping would be using two equal amplifiers. The main advantage of this way of operation is that the amplifiers have to deliver lower current, specially the one connected to the treble, and most amplifiers sound better when they do not have to deliver high currents.
Most speakers have complex crossovers and their transfer functions can not be emulated using the typical active crossovers, such as the Berhinger DCX2496 or Marchand's, that only allows the selection of a few basic filter shapes, slopes and phases. Should we expect that an approximate transfer function is better than the original, carefully studied and optimized by the designer?
Can you elaborate a little bit? I would have thought the contrary. From my pint of views The lowly DCX-2496 covers 3/4 of the most commonly used filter in speakers .. Bessel, Butterworth and Linkwitz-Riley .. I don't know (doesn't mean there aren't any) of a speaker with the very rare Tchecbyshev filters ...
The Behringer can provide all kind of slopes, as a matter of fact, slopes passive can't even begin to emulate so .. I am wondering. I am not vouching for its sonic qualities simply to the fact that I was under the impression that if you knew the type of filters used in a speaker you could emulate it with the Behringer on the active side ...
And there are other solutions such the DEQX , the TacT which are more "audiophile", others are using the Fab Filters as Eq and /or Crossovers .. There are countless Digital filters, that can emulate anything analog you throw at them ...
There are other even cheaper solutions on the active filters side the min-DSP comes to mind , I don;t know how it sounds, I honestly have my doubts on its sound quality but in term of flexibility not much you can't do with this little thing...
I believe one person besides Gary Protein, here Keith W (?) has done exactly that. he emulated the crossover in his speakers with a DEQX and built an active filter or had it built, based on the results from the DEQX whose sonics he din't quite like. He can always chime in ...
Can you elaborate a little bit? I would have thought the contrary. From my pint of views The lowly DCX-2496 covers 3/4 of the most commonly used filter in speakers .. Bessel, Butterworth and Linkwitz-Riley .. I don't know (doesn't mean there aren't any) of a speaker with the very rare Tchecbyshev filters ...
The Behringer can provide all kind of slopes, as a matter of fact, slopes passive can't even begin to emulate so .. I am wondering. I am not vouching for its sonic qualities simply to the fact that I was under the impression that if you knew the type of filters used in a speaker you could emulate it with the Behringer on the active side ... (...)
There are several aspects to consider. Just look at the schematic of the crossover of the Dynaudio Ayres, a simple 2 way speaker. Do you think that you can easily emulate the transfer function of the treble section?
Or if you still do not believe look for the techniques developed by Harman to interactively design the crossovers of their speakers - do you believe that if all it was needed were just some ideal Linkwitz-Riley, Bessel or Butterworth filters they would spend so many resources optimizing its design? Also you have to consider that the speakers themselves have an impedance that is not purely resistive - this means that you have also to consider its variation versus frequency and use a complex model of the speaker impedance. One last aspect that I am thinking about is the variation of the phase of the crossover with frequency - you can not be sure that you respecting it with an active crossover.
I can easily accept that you can built fantastic active speakers, I am not so sure about fiddling with an existing optimized passive design.
This does not prevent a few people from getting excellent results with this technique - but these are particular cases of dedicated people with a lot of expertise.
