|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.109.157.130
In Reply to: Re: All right... posted by BrassTeacher on February 21, 2007 at 20:50:11:
The component count in a piece of gear is a completely bogus argument--your preamp, power amp, CD player and tuner contain far more components than any electronic crossover. Does that mean you should eliminate them?Obviously not. The point here is that the RC circuits of a quality electronic xover contain much smaller value capacitors (note that there are typically NO inductors), as compared to the much larger value components in a typical passive xover network. The gain stages of a typical electronic xover (IC op-amp or tube-based) do a much better job of driving their associated small-value xover networks than the typical power amp does driving its large-value passive xover. Better transient response, fewer unpredictable phase shifts and response anomalies, etc. In a nutshell, better sound.
Yes, this is all predicated on driving the individual speaker drivers directly, something I neglected to mention before (in the interest of brevity), but which should be obvious. Yes, bypassing the passive xover in a multi-driver dynamic speaker such as the brands you mention is a risky proposition, for the reasons you mention, and I probably wouldn't attempt it myself. Active bi-amping with an electric xover is much more successful with planar speakers that don't need response tailoring, such as Newform Research (which I own) or Magnapan, as somebody has already pointed out.
Follow Ups:
I think you must pay attention to the inductances of a passive xover. You can find easily ohmic values of tens of ohm. If you have, for example, a 4 ohm speaker with a 0.1 ohm resistor placed in series, you will get a DF worse of 4/0.1=40 whenever amp you use, so with a passive crossover let's forget the DF=4000 of Nuforce...
Ops...I meant few tens of mohm (milli-ohm) not tens of ohm...
Hi C.B.I don't think the component count point is bogus at all. In a way, you kind of make my point for me. If using all those 'passive' components is such a bad thing, my point is that even an active crossover adds many passive components to the signal path, smaller part values or no. Also, getting rid of the tuner, CD player, preamp, poweramp is a moot point, you must at least have a source, and something to apply enough gain to that source to drive some sort of transducer.
My way of thinking is this, the simpler something can be, the better. I personally don't do this, but a lot of people DO eliminate their preamp, simply to get as many things as possible out of the signal path, and simply use a selector switch, and some kind of attenuator as a "preamp" (I'm sure you know this, but not everyone reading might).
And, actually, my preamp (my design) only has 1 opamp (well, 2, it's a dual), 10 capacitors (would be less if some weren't bypassed), 5 resistors - and that's counting the volume pot, and 1 selector switch. I didn't include power supply parts in the active XO count, so I leave those out of this count. They are about the same anyway.
Why so few resistors? To me, resistor noise is MUCH easier to hear than when compared to just about any other component. So, back to our active crossover, which added SIX more potentiometers to the mix, and by the looks of each pot's function, at least 4 of those, probably 5, are directly in the signal path. That worries me more than an inductor. In my entire system, front to back, there's only one potentiometer, period. Also, that one pot is definitely of the "not suck" variety, I just hope my wife doesn't find out how much it cost! :)
I've taken similar steps to make sure there are as few items in the signal path as can safely be done away with in other parts of my system as well, but I need to learn more about power amp design to really be able to effectively un-clutter things.
Now back to our active crossover - who's to say the designer of said active crossover wasn't paying attention, and part of their circuit inverts its part of the signal? I'm not talking about phase shift, I'm talking totally inverted polarity. It evidently is a fairly easy mistake to make, as it's my understanding that most CD players actually invert polarity. I recently found out mine does, when I aquired a preamp that has a polarity switch (It's a big enough difference I'm going back and adding a polarity invert switch to my normal preamp). The CD player isn't some thrown-together piece of junk either. It's older, yes, but in it's day it was very highly regarded, and it still holds its own compared to a lot of modern offerings.
Also, if you need the gain of an active XO to get the transient response you need, I would suggest there's something REALLY wrong upstream.
Then again, you have made me curious enough to at least look into it. Now I wish I still had that pair of Magnepan MG IIIa's. They would've been a lot easier to experiment with.
Now, the active XO design I would want to try, does one exist that:
1. has unity gain.
2. crossover point, slope, and relative volume level to the other drivers is set by individual fixed resistors and caps in the circuit that can be adjusted by plugging in, such as can be done with a lot of phono preamps to tailor the response of the cartridge to the preamp.
3. that the circuit designer not only closely paid attention to phase (time alignment), but making sure no part of the signal has its polarity inverted.
4. has all of the above.
5. has all of the above with the most efficient number of components as possible.Hmmm, now you've got what's left of my brain churning...
...lemme go find my stack of opamp data sheets, if someone already builds something like the above, let me know, otherwise, I just might build one...
...anyone in the Atlanta area got a spare pair of Maggies I can play with?
BT
This post is made possible by the generous support of people like you and our sponsors: