Quote from: Rich B on 2015-06-02 20:33:07

An engineering fellow told me that the separate amps in the AVR may run cooler, or that there may be a better and more even distribution of heat.

What would that have to do with sound?
I've already linked you to the only possible benefit, full clipping spectra or low passed clipping spectra. Key being "clipping", you would have to drive the amplifier into some form of non-linearity. Which is possible.

Quote from: Arnold B. Krueger on 2015-06-02 20:32:41

I would say nothing that is worth the trouble, but enough for someone that we both know to make a big thing about it on AVS. ;-)

You referring to jj's thoughts on the subject? He would tell RichB the same thing I did. Depends.

cheers,

AJ

My engineering friend read the thread but nods his head. He says, and I'm paraphrasing : bi-amping into loudspeaker cross-over filters, means that when one amplifier is working the other one is not. It is the same as connecting the amplifiers in parallel (theoretically), thus in essence having one amplifier only. There is however the point of headroom. When the peaks of a l.f. and h.f. signal happened to occur simultaneously, the voltages add, which will not happen with separate amplifiers. Thus a bi-amping system will have a higher headroom capability.

He says, and I'm paraphrasing : bi-amping into loudspeaker cross-over filters, means that when one amplifier is working the other one is not. It is the same as connecting the amplifiers in parallel (theoretically), thus in essence having one amplifier only. There is however the point of headroom. When the peaks of a l.f. and h.f. signal happened to occur simultaneously, the voltages add, which will not happen with separate amplifiers. Thus a bi-amping system will have a higher headroom capability.

Quote from: Rich B on 2015-06-03 06:40:10

He says, and I'm paraphrasing : bi-amping into loudspeaker cross-over filters, means that when one amplifier is working the other one is not. It is the same as connecting the amplifiers in parallel (theoretically), thus in essence having one amplifier only. There is however the point of headroom. When the peaks of a l.f. and h.f. signal happened to occur simultaneously, the voltages add, which will not happen with separate amplifiers. Thus a bi-amping system will have a higher headroom capability.

Both amplifiers are working and outputting the same voltage, the same fullrange signal... As for current and power, well, the LF amp will maybe have to deliver a dB less because instead of the tweeter it will see the increasing impedance of the low-pass filter.
It's only 1-2 dB or so because the spectrum of music behaves a bit pink noise so power falls off about 3 dB per octave.

The HF amp needs to output a lot less power, but needs to be running so overall efficiency will be (quite a bit) lower.


Clipping due to running out of voltage swing will be identical in both amps, again because both amplify the same signal. Only in terms of power you can gain the aforementioned dB.


This is why it's called fool's bi-amping, because if you run into clipping you should simply buy a better amp that will give you a lot more headroom. In case of an AVR I see no problem, but also little advantages.

Both amplifiers are working and outputting the same voltage, the same fullrange signal... As for current and power, well, the LF amp will maybe have to deliver a dB less because instead of the tweeter it will see the increasing impedance of the low-pass filter.
It's only 1-2 dB or so because the spectrum of music behaves a bit pink noise so power falls off about 3 dB per octave.

Quote

Both amplifiers are working and outputting the same voltage, the same fullrange signal... As for current and power, well, the LF amp will maybe have to deliver a dB less because instead of the tweeter it will see the increasing impedance of the low-pass filter.
It's only 1-2 dB or so because the spectrum of music behaves a bit pink noise so power falls off about 3 dB per octave.

So if both amps are outputting the same voltage and signal then there is a theoretical advantage of 3 db.

Am I on the right track?

So if both amps are outputting the same voltage and signal then there is a theoretical advantage of 3 db. But the tweeter won't make use of the extra power because it doesn't need it and hence it is a waste of power = worse efficiency?  So only the woofer could benefit in such a scenario?

And the only way to make use of this theoretical benefit is if there was sufficient voltage swing from the amp and assuming the speaker could handle the extra output without thermal compression taking place? Am I on the right track?

No there is an advantage of less than 1 dB and then only if you driving the amp to clipping all of the time, which is is neither realistic nor significant.

In real world use, as you have been told several times already Rich, there is zero, zip, nine, nichts, nada significant real world benefit.

Quote from: Arnold B. Krueger on 2015-06-03 14:58:03

In real world use, as you have been told several times already Rich, there is zero, zip, nine, nichts, nada significant real world benefit.

Well, it depends on what you mean with significant.

A dB can make an audible difference in corner cases.

Quote

No there is an advantage of less than 1 dB and then only if you driving the amp to clipping all of the time, which is is neither realistic nor significant.

But how did you calculate less than 1 dB??? It seems people are saying there is no headroom benefit so then how do you reach that conclusion?

I did a simulation with speaker models found here:
Design of Passive Crossovers
using the "2dB L-Pad Attenuator" crossover network but without the zobel correction (which would just turn the power 'advantage' into heat).

A logarithmic 1s long sweep from 20Hz - 20kHz results in:
1.3 dB less power output from the LF amp
5.6 dB less power output from the HF amp

This is with 100% efficiency and no losses.

I estimated  the change in voltage dropped across the output stage due to offloading it by putting the tweeter on another amp.  There will be a finite decrease, but it will be pretty small.  Note that zero is a possible value for "less than 1 dB".

Let's pretend ... "Wow, look how much louder I can [effectively] now play the music thanks to passive bi-amping" is extremely narrow* before you hit a brand new, but different road block.    When the tops of the waveforms start to flatten, the onset of clipping, not only have you reached the limit of that amp's clean sounding range you also have reached the end of its music amplification. Going any louder with the main preamp's volume knob will increase the tweeter amp's output, but all that happens with the already distorting woofer amp is an even flatter, nearing square wave shape to its waveform. It doesn't play the music more loudly like it should. Your overall system's highs are successfully increasing in level but your lows are maxed out so the spectral balance of the music will become too bright (treble heavy/bass shy).  So how much louder can you play a system before you notice the tweeters are playing at a higher level than they should be, compared to the woofers? I'm thinking 1 dB, maybe even less with the right material.

Arny, please help me out here. Can you please cite for us ABX studies for level discrimination which inserted time delays of seconds or better still minutes between evaluations? Acoustical memory for exact level is fleeting.

Arny, since you own oscilloscopes and line out converters, I don't, would you please confirm or deny that say a 1 kHz  test tone itself [not its spurious distortion signal] at the point that it clips coming out of a power amp no longer gets any louder as you continue to increase the incoming preamp signal even more. All that happens is you get lots of extra distortion and other grunginess but the top of that waveform's amplitude pretty much stays put (or at least starts to become severely compressed from where it should be), as it gets flatter and flatter, but you cant say the amplitude is correctly tracking the input signal level anymore. Correct?