each channel may be seeing 16 ohms nominal This was in my first post and in error. I was thinking the highs are easier to drive and may see less power when on a separate amp. |
Yes, indeed, Al, I was ref'ing the amount of POWER needed by frequency band. The 50:50 point is around 350hz while 4khz means you need only a fraction of the power total at and above thate frequency.
That's why people who have a pair of identical mono's set as biamp with the crossover at higher frequencies are wasting the potential of the HF amp.
In my fantasy world, I'd want BOTH amps to redline at the same point.
Here is link to a fine bi-amping article, which if I failed to link is my bad.
http://sound.westhost.com/bi-amp.htm
My panels cross at 600hz, so while the ratio is not quite 50:50, I wouldn't feel too bad getting another duplicate amp for the bi-amp attempt. I'm guessing that a kilowatt@4 per speaker should about settle the 'enough power' question for Magnepan 1.6s or maybe not!
I'd like to see measured data for the B&W685. If it follows the pattern of many other B&W systems, I woudn't attempt the Tube Solution. |
11-17-11: Hifihvn
One odd thing that I can never figure out is, a knowledgeable person that may be a manufacturer (speaker?) that posts (or used to) here sometimes said that when you bi-amp, the load is divided equally on the speaker inputs when their post separated. This was in the past few years. I don't recall who it was, but that never quite made sense to me. I looked at some old forums and can't find it, and may be gone. Oh well. I do think the impedance can be divided more closely between the highs and lows on a close to linear speaker like a Magneplanar though. I suspect that he would have been referring to an approximately equal division of the amount of POWER that would have to be delivered by the two amplifiers (or amplifier sections), which I would expect to occur with typical music if the speaker had a crossover point between the two sections somewhere in the low 100's of Hz. I would not expect that to be applicable to speakers that don't have a separate mid-range driver (and even to many that do), because in most or all of those cases the crossover point would be much too high to result in an equal division of power. Best regards, -- Al |
Of course, a lot of this is over my head, but I really can't see (due to ignorance?) how the load can be divided equally. I only did that in my model for simplicity's sake.
What leads me to believe that the loads are not equal is my one experience with active speakers. In my case, it was a pair of Paradigm Studio 20s. If I remember right, the highs had a 50 or 75W amp on them and the lows had a 100 or 125W amp on them.
What I've gathered from the information that you guys have provided so far is that getting the better sound is about the lack of distortion that is brought about by not having to stress the amp.
What I'm hoping all of this is telling me is that if I get an amp with some nuts, I won't have to worry about bi-amping in the future. |
Magfan, I haven't owned a B&W since they went toward metal dome tweeters, and I don't keep up with them to much anymore. I'm just thinking it is hard to put a set figure with too many variables. These three speakers will give totally different results on a tube amp vs. a solid state in my opinion. The OP's amp is a 5 channel, and probably can't give a linear output across all frequencies like a potent solid state one or two channel can. I do remember the killer Apogees and Infinity Kappas.
One odd thing that I can never figure out is, a knowledgeable person that may be a manufacturer (speaker?) that posts (or used to) here sometimes said that when you bi-amp, the load is divided equally on the speaker inputs when their post separated. This was in the past few years. I don't recall who it was, but that never quite made sense to me. I looked at some old forums and can't find it, and may be gone. Oh well. I do think the impedance can be divided more closely between the highs and lows on a close to linear speaker like a Magneplanar though. |
IF the '685s follow normal B&W practice, they are a real handful in the lowest couple octaves and may even dip to 3 ohms with a largish phase angle. This is amp killer territory. With a crossover at 4khz, you could do with an amp of 1/4th the 'wattage' above crossover than below....easily, and never clip the HF amp.
http://sound.westhost.com/bi-amp.htm
It wouldn't surprise me even a little that the combination of being able to turn it up, slightly, and the reduction in HF distortion allowed a higher playback level with the same perceived loudness as 'before', while the bass balance shifted lower. Same SPL to within a few db. Even on my system with LOTS of power, the bass opens up when I play at a radioshack analogue meter'd 80db. I believe ear sensitivity plays a large part in this...at lower levels, especially.
The impedance curve of the speaker? almost irrelevant when dealing with a SS amp. Note from the Stereophile test of the Harbeth you link that potential probems with rising response were with a Tube amp...not SS. Stereophile is generally more concerned with wacky phase data when it occurs WITH impedance minima. |
Also, with the load shifting toward the low end, the OP said his "bass tightened up and went lower". |
11-17-11: Magfan
With a 4khz crossover, at least 80% of the power required for 'normal' music is below that frequency.
I think it varies a lot depending on the speakers impedance curve, besides the music, and crossover point. Every speaker amp and speaker combination may have different results. That's why I can see what the OP results were. It would be nice to have this type of test results for all speakers. Some examples from Sterephile. [http://www.stereophile.com/content/magnepan-magneplanar-mg36r-loudspeaker-measurements] [http://www.stereophile.com/content/atc-scm-11-loudspeaker-measurements] [http://www.stereophile.com/content/harbeth-p3esr-loudspeaker-measurements] |
Sometimes distortion is perceived as loudness. There is a possibility that while driving your speakers with more power you have lowered the distortion level of the speakers enabling you to increase the volume without negative affects. |
With a 4khz crossover, at least 80% of the power required for 'normal' music is below that frequency.
So, you are still working the lowfrequency amp pretty hard, but not quite as hard as before. You have the addtional beneift of the hi frequency amp simply loafing along. Nothing you could do would tap THAT part of the amp out, so your distortion at higher frequencies, where fundamentals leave off and the overtones predominate, is pretty much history. I suspect that is the source of the help you perceive.
The low impedance dip is a little bothersome, especially when driven by HT amplifiers, which are notoriously unfriendly to such specs. |
If I'm looking at the right amp on the net, the PA-5800 is a 5 channel. If your using 4 of these channels, I can see the volume and wattage drop. This would be due to a lighter load on each channel of the amp. I believe each channel will be putting out less in this case. Just guessing, each channel may be seeing 16 ohms nominal, and causing less power due to the lesser load on them. Anywhere from 40 to less than 80 watts per channel, except when your speakers drop to that lower impedance at certain frequencies. The sound can be different from this lighter load. [http://www.harmankardon.com/resources/Brands/harmankardon/Products/ProductRelatedDocuments/en-US/OwnersManual/PA5800%20om.pdf] |
Al, I really appreciate your taking the time. Like an idiot, I completely forgot about the crossover and the affect that it would have on the load, or its distribution. You really cleared things up for me and now I almost regret having posted. I could have answered my own question (or gotten pretty close to it) with a little more digging.
As for the volume level, it could be all in my head. That tends to happen to me sometimes.
In the end, I do know that MY system sounds better in this configuration and would have accepted the trade off in power (had there been one) for the better sound. |
I'll assume that each component in the speaker is running in parallel to the other. If that is right, then (to keep things simple) can I say that each component has a resistance of 16 ohms so that when I run the components in parallel I have a speaker that presents an 8 ohm load?
If all of this is right, then by bi-amping, I've effectively made it so that each channel of my amp is driving a 16 ohm load. Again, if this is true, then I've taken my 80 wpc amp and turned it into something around a 50 or 60 wpc amp, due to the greater load. No, the low and high frequency sections of the speaker are not in parallel, due to the presence of the crossover network in the speaker. The amplifier section that is driving the speaker's woofer will NOMINALLY see an 8 ohm load for the very low frequency components of the signal, and a very high impedance for the very high frequency components of the signal. Therefore that amplifier section will have to supply little or no current and power at very high frequencies. At frequencies in the crossover region, which for the 685 is specified as being centered at 4 kHz but is probably fairly broad, the impedance seen by that amplifier section will gradually transition between the nominal 8 ohm value and being very high. The opposite situation will exist for the amplifier section driving the tweeter, i.e., it will have to supply little or no current and power at low frequencies, and it will see a nominally 8 ohm load at high frequencies, supplying whatever amount of current and power the signal calls for at high frequencies. The actual values of the 8 ohm nominal loads will vary depending on how the speaker's impedance varies as a function of frequency. I note that the impedance of the 685 is specified as 8 ohms nominal, 3.7 ohms minimum. I would suspect that the 3.7 ohms occurs at frequencies that are handled by the woofer, which are also the frequencies that almost always require the most power. I don't know why you might be finding that you have to turn the volume control to higher settings in the biamp configuration, unless the gain of the amplifier somehow changes when it is set up for four channel operation. Regards, -- Al |
