|
Post by Boomzilla on May 31, 2013 12:40:06 GMT -5
...how much of the back-EMF signal actually reaches the tweeter in a parallel crossover? The correct answer is zero. Should be so.
|
|
KeithL
Administrator
Posts: 10,147
|
Post by KeithL on May 31, 2013 13:24:47 GMT -5
Wrong. The reason is NOT about unequal lengths causing unequal delays. (You are quite right there; the delay is insignificant.) Here's the reason: ALL cables have electrical characteristics; they all have resistance, inductance, and capacitance - and all of these are measured "per foot" (a longer cable has higher resistance, and usually higher capacitance and inductance than a shorter one made of the same type of wire). Even though most of the claims about particular specific values being critical is hogwash, these characteristics *do* alter the sound slightly. As such, you want both wires in each pair (both fronts, both surrounds, etc) to be the same so that the signal reaching both speakers will be the same. [Within very wide limits, there is no real right or wrong, but you DO want them to be THE SAME. A twenty foot cable of a certain type WILL have twice the resistance of a ten foot cable of the same type. Either one is fine within acceptable limits, but they will be measurably different.] The resistance of ten or twenty feet of speaker cable is NOT insignificant in all situations. It can easily reach a few tenths of an ohm which, with can interact with the impedance of certain speakers to produce variations of up to a few tenths of a dB (or even more) in frequency response - which may be audible. Even though these variations are tiny in absolute terms, differences between left and right of that magnitude can be audible (and, yes, most speakers aren't matched that carefully anyway - but some are). We're not talking about power losses; we're talking about the R in the cable interacting with the R, L, and C in the speaker and acting as a filter. I'm not suggesting that it's critical to measure to the inch, but since I can't think of a good reason NOT to make the wires the same length, it couldn't hurt. Regardless of whether it's a front, side or rear speaker pair, both speakers in a speaker pair (Left and Right) should always have identical cables with identical lengths. That is conventional wisdom however it really doesn't matter. Since audio signals travel through copper wire at approximately 60% the speed of light, it would take a length difference of many hundreds of feet before even a few milliseconds of delay is introduced. And unless you are using small wire gauge, the resistance differences of less than a few dozen feet create a completely negligible impact on power differences. Get them as close as is practical, not a bad idea, but the idea that lengths must be exactly the same or some bad effect results is simply not true.
|
|
|
Post by jmilton on May 31, 2013 13:35:37 GMT -5
|
|
DYohn
Emo VIPs
Posts: 18,439
|
Post by DYohn on May 31, 2013 13:55:07 GMT -5
Keith, like I said, if you are using small gauge speaker wire then yes, length differences can introduce unwanted resistance effects, including but not limited to power differences and in very severe cases RLC filter effects. But I maintain it is not "Wrong" to suggest that unequal lengths have next to no impact ASSUMING the wire is of sufficient gauge. Which I stated.
|
|
|
Post by Boomzilla on May 31, 2013 16:50:19 GMT -5
With my 8-ga. "Godzilla" wires that I made myself, I doubt that a 3-foot vs. a 12-foot run would make ANY significant difference at all in series resistance. Capacitance & inductance, maybe (or maybe not). Since I made both pairs 10-feet long, I'm not about to butcher them just to see if I can hear a difference.
On the other hand, next time I make a pair (I'm prone to such lunacy), I may buy enough spare wire to try a "shortie" vs. a "longie" and just see...
|
|
|
Post by yves on Jun 1, 2013 4:20:36 GMT -5
Indeed, in a bi-wire situation you are actually INCREASING the impedance between the speaker and the amplifier by adding speaker wire, which lowers the damping factor. A jumper adds much less impedance to the system and is, in this respect, better. A speaker that properly supports bi-wiring has two separate crossover networks. You can do the 1.5V battery trick Bob Katz mentioned in that Gearslutz forum thread I linked in my previous reply to this thread. Standard bi-wiring doesn't increase the impedance between the speaker and the amplifier at all. This is because if the crossover networks are not connected to eachother anywhere except at the binding posts of the amp, the signal created as a result of back-EMF will travel back to the amp via only one pair of wires (i.e., one positive and one negative wire placed together as a pair inside only a single length of cable), and this obviously holds true no matter which one of both crossover networks you're looking at, as you can easily see by just looking at the diagram shown below.
|
|
|
Post by yves on Jun 1, 2013 8:12:49 GMT -5
What is the frequency of woofer back-EMF? The answer is it is exactly the same as the woofer signal, only with reversed polarity. The signal produced by electromagnetic inductance as a result of the woofer's back-EMF isn't the same as the woofer signal with inverse polarity. The woofer signal, produced by the power amp combined with the speaker cable combined with the low frequency crossover network, is what causes the woofer to vibrate. The woofer then continues to vibrate, even after the woofer signal has long disappeared from the chain and, as a result of this continued vibration, the back-EMF continues to be present regardless of the woofer signal. The power amp effectively receives the signal this back-EMF creates, and adjusts its own output signal accordingly in response, in order to damp the woofer's continued vibration. In a bi-wired scheme, the signal caused by the woofer's back-EMF travels first from the low frequency crossover network back to the amp via two wires (i.e., one positive wire and one negative wire placed together as a pair, to form a signle length of cable) and travels next to the high frequency crossover network via two more wires (i.e., one more positive wire and one more negative wire placed together as another pair, to form yet another single length of cable). ['First' and 'next' here meaning the two lengths of cable, as they connect both crossover networks to eachother, are now effectively in series rather than in parallel, at least from this particular point of view...]. So, if moving from mono-wiring to bi-wiring, one wire hooked up in series with a second wire replaces one of both removable links, and a third wire hooked up in series with a fourth wire replaces the second one of both removable links. That's two pairs of (usually) relatively long wires versus only one pair of relatively very short removable links, per speaker. [However, like I said in my previous reply, this will not increase the impedance between the amp and speaker...]. Anyway, let's not ignore the electromagnetic properties of the insulator(s), or dielectrics, combined with the effects of electromagnetic inductance. This inductance causes signals to electromagnetically interfere with other, nearby signals (if any). In the mono-wiring scenario, the signal resulting from the woofer's back-EMF can travel through the speaker cable in one direction only (i.e., from the speaker toward the amp) because the only other path available to it is through the removable links. Bi-wiring cables (such as the ones that I have) have four wires inside them rather than just two, which will cause this same signal to make a U-turn at the amp's binding posts and thereby utilize the two extra wires inside the same cable to travel back toward the speaker whence it came from (albeit toward the other one of both of its crossover networks, rather than toward the same crossover network whence it came from). Therefore, by means of oversimplification, it can be seen that half of this signal's U-shaped path gets surrounded by half of an electromagnetic field that, in a polar-inverse type fashion, partly overlaps the other half of this same electromagnetic field that surrounds the other half of this same path (i.e., we're dealing with a conductor that can technically be also called an inductor). Finally, note that this signal path is in fact not really a U-shape, but a U-bolt because the amp is connected to the bottom of the U-shape. [Had this not been the case, I guess you could argue that mono-wiring, if using removable links that are both U-shaped as well and sufficiently long, can be the electric equivalent of bi-wiring...] The signal running through the U-bolt is actually the sum of (A) the amp's output signal that goes from the bottom up via both sides of the U-bolt, (B) the signal that comes out of the low frequency crossover network and enters via the top of one of both sides of the U-bolt, (C) the signal that comes out of the high frequency crossover network and enters via the top of the other side of the U-bolt, (D) the signal caused by electromagnetic inductance in response to fluctuating electromagnetic fields that surround the U-bolt, and (E) the signal caused by various 'negligible' (or should I say, possibly audible...) other effects, such as skin effect and proximity effect. Simply put, the reactance part of complex impedance is typically making circuit analysis a little more complicated than you normally would expect. [Some say skin effect also can (or does) matter because it affects very high frequencies the energy of which, when it enters the rectifier of an amp, is turned into noise that covers the audible band of the spectrum, and that such differences can indeed be measured, albeit not if using ordinary types of measuring devices].
|
|
DYohn
Emo VIPs
Posts: 18,439
|
Post by DYohn on Jun 1, 2013 10:06:18 GMT -5
Yves, you believe whatever you need to in order to justify your decision. But the fact remains that in a parallel crossover network exactly as you have depicted, the effect of any back-EMF from the woofer is moot. Plus, whoever you are cutting and pasting in the second post above - or if those are your words it's unclear - does not fully understand how a dynamic loudspeaker works. A woofer does not "continue to vibrate" except in response to a signal. The signal controls the woofer cone movement very precisely. Otherwise the sound would be terrible. Any non-signal related movement (except for the cone returning to it's "home" or "equilibrium position" when all signal is removed) is a form of distortion and is minimized in a properly designed woofer by a compliant suspension system, and usually designed-around in a properly designed Xover network. Signals generated by this distortion are measurable of course but they really have no effect on crossovers or crossover design. They are sometimes audible because they also generate sound that can be heard, but the impact is not improved by bi-wiring. Indeed it is exactly the same no matter how one wires the system as long as the crossovers are connected together at some point as they must be. There is no electrical difference between bi-wiring and single-wiring or quad wiring or even using wireless connectivity. It all works the same way.
But like I said, you keep enjoying your setup, keep believing that you have done something wonderful, and I am sure you will find it sounds wonderful as well. Our brains are powerful tools.
|
|
emovac
Emo VIPs
Saeed al-Sahhaf
Posts: 2,456
|
Post by emovac on Jun 1, 2013 11:33:45 GMT -5
+1 on Blue Jeans Cables. Also check out the Canare 4S11 if you plan a bi-wiring setup.
|
|
|
Post by flamingeye on Jun 2, 2013 10:26:47 GMT -5
monoprice 12awg use the rest of your money on good recordings
|
|