Passive Bi-amping

[DonH50]

Funny, being totally unfamiliar with the pro-world, bi-amping is using 2 amps to drive a passive speaker for me. When reading an article by a pro about bi-amping it always take some time before I realize this is about using an active cross-over.

Likewise the phrase “passive bi-amping” is new to me. But very apt.

Any thought on a typical audiophile issue as vertical or horizontal (passive) be-amping?

Might it be that there is a slight improvement because driving 1 speaker over a low or high pass filter is a less complex load for the amp than driving 2 speakers over a high + low pass filter

I have to check the context now whenever I see anything about "bi-amping".

Horizontal vs. Vertical amps: If you use one amp per side you have better channel isolation and any power modulation is "all together" with the signal. You can also place the amp physically close to the speaker and use very short cables from amp to speaker. If you use one amp to power bass channels and one to power highs, there would be less modulation of highs by lows (and vice versa) through the power supply (etc.) In an active system you can also use a smaller amp for the highs. I have done it both ways and am not sure there's a clear winner, but sonically most (including me) prefer to use a big honkin' amp on the bass and smaller, "smoother" amp on the highs. The most popular combo (in the audiophile world) that comes to mind is a SS or hybrid bass amp and tube mid/upper amp. That is what I used to do with my Maggies.

It is possible the load is "easier" on the respective amp channels due to separating the low and mid/high impedance variations. That would be quite speaker- and amplifier-dependent, natch.

[DonH50]

I think in the majority of multi-amp systems there would always be a level of local bypassing in every amplifier module. My experience is that high frequency glitches in supplies do the most damage, and of course a capacitor in exactly the right position in a circuit is worth 10 further away in the main power supply bypassing, as far as high frequencies are concerned. Many amplifiers have great PSRR, so long as you don't check how they perform at those nasty upper frequencies!

Frank

True, PSRR and CMRR could be better at high frequencies (where I agree they tend to fall). As I said, I have always used separate amps in the sound reinforcement arena...

[DonH50]

Don it would be great if, some time, you could do a breakdown like this for active biamped systems. I have come across some explanation of the benefits I clearly hear, but none as clear and methodical as the above.

Tim

Active bi-amping uses a line-level crossover (which can be active or passive) to separate the frequency bands before the power amps. This allows the crossovers in the speakers to be bypassed and the amps to be connected directly to the drivers. Individual amplifiers can be optimized for each frequency band. A more powerful amp is often used for the bass, and a lower-power design for the mid/high range (which typically requires much less power). This also allows users to choose the “flavor”” of amp desired, e.g. SS class AB (or D) for tight bass and lots of power in the low-range amp, and a tube or hybrid design, perhaps class A, for the upper frequencies.

Some points to consider:

1. Amplifier power can be optimized for the frequency band, providing appropriate headroom without wasting energy. There is no more power gain than with a passive system, but power can be distributed more appropriately since each amp carries only a limited frequency band instead of running full bandwidth.
2. Without the passive speaker crossover in series between the amp and speaker (driver), driver control (damping factor) is improved, power loss in the crossover is eliminated, and impedance is generally more predictable/controlled.
3. You do need more amplifiers, which means more space, potentially more waste power and heat, more cables/interconnects, and of course you need a crossover (again, line-level, active or passive). Matching the “sound” of the amps to each other and to the speaker can be a challenge; you probably do not want a drastically different sound in the two frequency bands (a SET mid/high amp and high-wattage SS monoblock does not strike me as a good match, though I am sure it has been done; it also depends on the speaker design, of course).
4. The crossover (typically) provides great flexibility, allowing users to select and optimize crossover frequencies, slopes, phase, level, etc. Without the trade-offs of high-power passive components the speaker design can be optimized and interaction among drivers minimized. High-order slopes are relatively easy to realize, and an active line-level crossover adds buffer amps and enables fine control of crossover parameters not possible with a fixed in-speaker crossover. This implies having the knowledge and equipment to adjust and optimize the crossover…
5. There is no electrical interaction among drivers with active (or passive) bi-amping. (There may still be mechanical coupling if the drivers are not isolated from each other.) That is, if the woofer starts to distort the input signal through electromechanical forces, it no longer modulates the HF amp’s output; they are separate amps.
6. If the amps share a power supply, then modulation among channels can still occur through the power supply. If discrete monoblocks are used for all channels, this is a non-issue. If stereo amps are used, you must choose to use an amp per side or one amp for lows and another for highs. A single amp per side pretty much eliminates crosstalk between L/R channels, but since you’ll have to size the amp for the largest signal (bass), one channel will be run very lightly. You also lose the ability to select high and low amps for “taste”. You gain the ability to use very short speaker cables by placing the amps near the speakers. Using separate high and low amps allows power optimization and selection for sonic “taste” in each band and is probably the more common approach.
7. Again, there may be some distortion reduction since the amps are not required to handle full-range signals (and drivers). And again, I suspect this is a minor matter. However, it may be arguably better with active bi-amping since the amps no longer need to handle the entire frequency band and signal levels, eliminating some of the potential for distortion caused by high-level lows modulating low-level highs.
8. You again have two amps now so presumably noise is a little higher since you have two uncorrelated noise sources. At the speaker outputs I suspect it’s a wash since only a reduced frequency band gets through the drivers to hear.
9. Thermally, active bi-amping allows you to utilize a lower-power mid/treble amp, so power loss and thermal issues (heat) are less a concern than with passive bi-amping, which almost requires matched amplifiers (since all amps handle the full signal band).

To my mind, the ability to refine the crossover network and eliminate the passive crossovers are the biggest benefits to active bi-amping, with optimizing the amplifier choices close behind.

As always, FWIWFM - Don

[Robh3606]

The crossover (typically) provides great flexibility, allowing users to select and optimize crossover frequencies, slopes, phase, level, etc. Without the trade-offs of high-power passive components the speaker design can be optimized and interaction among drivers minimized. High-order slopes are relatively easy to realize, and an active line-level crossover adds buffer amps and enables fine control of crossover parameters not possible with a fixed in-speaker crossover. This implies having the knowledge and equipment to adjust and optimize the crossover…

Hello Don

Are speaking about Digital Speaker Controlers in this case. The only analog crossovers I am familiar with where you can readily change out crossover slopes and frequencies are either card based or jumper/switch bassed and I don't think any of them tackle phase. If there are analogs out there that have that kind of flexabillity could you please share them.

Thanks Rob

[microstrip]

Don,

As my speakers (electrostatic SoundLabs are not very efficient and have a complicated load - impedance varies between 90 and 1 ohm, with large phase angle variation) and allow bi-wiring, I have used them in by-amping for some time. The amplifiers were two conrad johnson premier 350 and there was a significant difference in sound between using a stereo amplifier or bi-amping with two amplifiers. Unhappily I never tried using just one channel of each stereo in a dual mono configuration and one should also consider I am not using a typical speaker.

It seems to me that a an immediate consequence of bi-amping will be that the currents flowing in each amplifiers will be lower, implying that power supply requirements for each amplifier will change and thermal effects due to power will also be lower. We also know that the distortion spectrum of amplifiers changes a lot with impedance - may the different loads that each amplifier sees in bi-amp mode also have a contribution.

[DonH50]

Hello Don

Are speaking about Digital Speaker Controlers in this case. The only analog crossovers I am familiar with where you can readily change out crossover slopes and frequencies are either card based or jumper/switch bassed and I don't think any of them tackle phase. If there are analogs out there that have that kind of flexabillity could you please share them.

Thanks Rob

Now, digital, yes. There are quite a few analog units that offer great flexibility in crossover frequency and gain settings, fewer with slope control, and I am not sure I have seen any with phase control (in days gone by I used an analog all-pass filter to change phase, and there were a few commercial units that offered that control). I don't have a list but did a search a few months ago. I ended up getting a donated dbx unit from mep that allows me to vary crossover frequency and gains but not slope; I did find units that offered everything but phase control but do not recall them offhand. It was a short list, and some were pretty expensive (for me, anyway). I'll (or you'll) have to dig up the threads in which I discussed my crossover search.

[DonH50]

Don,

As my speakers (electrostatic SoundLabs are not very efficient and have a complicated load - impedance varies between 90 and 1 ohm, with large phase angle variation) and allow bi-wiring, I have used them in by-amping for some time. The amplifiers were two conrad johnson premier 350 and there was a significant difference in sound between using a stereo amplifier or bi-amping with two amplifiers. Unhappily I never tried using just one channel of each stereo in a dual mono configuration and one should also consider I am not using a typical speaker.

It seems to me that a an immediate consequence of bi-amping will be that the currents flowing in each amplifiers will be lower, implying that power supply requirements for each amplifier will change and thermal effects due to power will also be lower. We also know that the distortion spectrum of amplifiers changes a lot with impedance - may the different loads that each amplifier sees in bi-amp mode also have a contribution.

Tube amps and difficult loads, I can see why even passive bi-amping would help. I consider that a special case, and a good one! I have done similar with various 'stats through the years, but always had a crossover before the power amps.

[JackD201]

My problem with active crossovers is that as they say, they aren't made equal. I used a slew of active crossovers in my days in sound reinforcement. Some were better than others and that's true for both analog and digital active XOs. While I can't deny the lack of efficiency in a passive bi-amp configuration, in my case where my loudspeakers have fully separate crossovers for the separate modules and these crossovers employ driver correction circuits, the ability to tailor my Midrange and Highs by using different amplifiers from medium powered tube amps to high powered solid state, gives me a degree of freedom I thoroughly enjoy.

Sure the DEQX can do driver correction as well but really, I'd have to move these behemoths to my friggin' driveway just to get the necessary conditions for proper measurements. It's fine for Albert, my speaker designer, whose personal system is a set of the same speakers run fully active with digital active XOs. He's got a huge factory to do proper measurements in.

To me it's not about power but subjective quality. Then again, I'm just a consumer not an engineer. My only responsibility is getting sound I like.