Zilch's AK Design Collaborative - Econowave Speaker
- Thread starter Zilch
- Start date
Let's put our collective minds to this:
http://www.audiokarma.org/forums/showthread.php?t=214088
What's the E'Wave answer?
[I have provided a clue over there.... :thmbsp: ]
1) I've never built a CC E'Wave, and,
2) ZilchLab is not a "critical listening" environment.
http://www.audiokarma.org/forums/showthread.php?t=214088
What's the E'Wave answer?
[I have provided a clue over there.... :thmbsp: ]
They're all over the place here, but:Zilch, have you heard a difference in a CC network? I am guessing you have tried it on an occasion or two...
1) I've never built a CC E'Wave, and,
2) ZilchLab is not a "critical listening" environment.
Paratima
E'Wavin'
The idea is to bias against zero-crossing distortion. This here CC stuff should be measurable, no?
Don't know if you'd catch this effect on a FR graph or RTA, although maybe on a waterfall. A dual trace scope hooked to the input and to the output of the CC'd network and drive a signal through it should show something, eh? We're not set up to see anything more than the secondary telltales of distortion, I think. :scratch2: Ya ain't gonna see this on a SPL meter from the Shack.
Don't know if you'd catch this effect on a FR graph or RTA, although maybe on a waterfall. A dual trace scope hooked to the input and to the output of the CC'd network and drive a signal through it should show something, eh? We're not set up to see anything more than the secondary telltales of distortion, I think. :scratch2: Ya ain't gonna see this on a SPL meter from the Shack.
Bauhausler
Rational Subjectivist
Ditto that. It's not the kind of experiment that I'd lend an ear to unless a measurement somewhere told me that the bias did linearize the capacitor's transfer function by pushing the zero crossing transition up to the peak of a 6.3VRMS signal where it's too tiny to affect anything.
If someone can find a white paper I'l lbe glad to read it, excerpt and summarize as best I can.
tubino
A guy with too many tubes
Thanks, that's exactly what I was asking.
So if I get B&C DE250s and adaptors to use with the $9.90 econo-JBL waveguide, can I still be an honorary EWG club member?
Indeed, many E'Wavers would likely be quite interested in how that combination might perform.... :yes:
Hello Zilch
There may be some modulation by the AC but it's still a DC source. You have the Time constant of the 2.2meg times the capacitance value to deal with as well. Just like you do in a conventional power supply. You are never going to drive the caps through 0 volts. You have a very clever and simple AC and DC circuit there. For AC it's a short through the caps and an open for the battery. For DC, once the caps are charged you have very little current draw on the battery and the capacitor discharge time is set by the resistor and capacitor value time constant. The discharge is not linear just like the initial charge time is not.
Rob
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html
The current available from the 2.2 meg source affords little competition to the low impedance AC signal. Even staying centered at the 9V operating point relies upon that signal being symmetrical, statistically:
http://www.audioheritage.org/vbulletin/showthread.php?p=130034&#post130034
Hello Zilch
Well let's agree to disagree. I don't see what that has to do with it?? The current available have nothing to do with the time constant. The time constant is one of the reasons it works. Once charged there is very little current required to keep the voltage quite high. Look at the discharge/charge curve. We are talking a music signal where peaks are very short duration. An 18 Volt peak to peak signal is going to be an issue on only the -9 to -18 volt portion with a + 9 volt DC source. That is the only portion of the curve where you could drive the capacitor through 0 volts. So 3/4 of the duty cycle there is no issue at all. That completely ignores the time constant set-up by the capacitor and the resistor. When are you going to have that type of voltage across the drivers where the duration is long enough to negate the time constant when it is not an issue for 3/4 of the duty cycle??
If that's your concern set them up with 18 volts ie 2 batteries. Easy enough to do.
Rob
I believe we agree, Rob, as to the practical outcome; I'm just suggesting that the time constant has little to do with it. It was clear when I looked at it that I would have to set up with dummy loads and a bigger amp to drive it through 0V. We're certainly not talking about anywhere near the requisite working voltages with compression drivers in home use. High-current woofers, maybe, but the caps aren't in line with those, typically....
Bauhausler
Rational Subjectivist
Hello Zilch
There may be some modulation by the AC but it's still a DC source. You have the Time constant of the 2.2meg times the capacitance value to deal with as well. Just like you do in a conventional power supply. You are never going to drive the caps through 0 volts. You have a very clever and simple AC and DC circuit there. For AC it's a short through the caps and an open for the battery. For DC, once the caps are charged you have very little current draw on the battery and the capacitor discharge time is set by the resistor and capacitor value time constant. The discharge is not linear just like the initial charge time is not.
Rob
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html
If one side of the cap is charged at +9VDC, then the dielectric of the cap will have exactly zero volts across it when the peak of the incoming AC signal equals +9V. This occurs with a steady state sine wave when the RMS value equals 9 / (sqrt)2 = ~ 9 X .707 = 6.36VAC. If there's a nonlinearity happening at the zero transition it would be swamped by the huge 6.36VAC signal, which represents approximately 5W of input to the driver in the case of a pass cap. If there's some sonic effect of the dielectric passing through the zero voltage point you definitely want that happening with a large input signal rather than with a very small signal where the nonlinearity comprises a large distortion.
I'm not saying this happens, just why you'd want to use bias that way.
It's like a class AB PP amplifier. You can design a Class B amp that has the amplifying devices on both phases of the signal 'turned off' and flowing no current when there's no ac signal to put out, but the transition from zero to current output is the amplifying devices' least linear point. The output impedance is high there, there's notch distortion, etc. It's easier to design a 'clean' output stage where the output devices are biased up with some overlap and both are flowing current at zero volts output. This keeps one of the output devices running in a fairly linear manner at all times.
If one wanted to 'hear' this effect I would recommend using cheap nonpolar electrolytics for the experiment. Something with a high dissipation factor. I would expect that adding bias to them would greatly improve their low level performance. That's not to say that the improvement would not exist with high quality film caps but it should be orders of magnitude smaller.
It's not that simple. A plus, plus a plus, is a plus so with a +9 volts + 9 volts peak you have 18 volts. You are completely ignoring the time constant which is significant. The only place you would approach zero volts is when the signal goes negative and only for the brief period of time where the voltage is -9 volts. At any other voltage above -9 volts and you are not at zero.
Hello Zilch
I don't think you took it into consideration and that is where we do disagree. You are also not addressing the duty cycle which increases the magnitude and influence of the time constant. The more time you have with the voltage above -9 volts the longer the caps will stay charged and not go through 0 volts.
Try a calculation to see what the time constant would be. Lets say you have a 1uf capacitor. To find T all you do is multiply Capacitance times Resistance. Multiply 1uf times 2.2 meg what do you get?? The discharge time to 99% is 5T so 5x that number for the caps to be 99% discharged. I come up with 2.2 seconds for 1T, take that out to 5T and you have 11 seconds. Try that calculation with a more typical 5 uf cap. Check my math but you can see where this is going.
You need a sustained voltage to bring them to zero that exceeds that time duration. You need 5T for a 99% charge or discharge. My point is it won't happen with music.
Rob
skywave-rider
Slumming in Akihabara
...
You need a sustained voltage to bring them to zero that exceeds that time duration. You need 5T for a 99% charge or discharge. My point is it won't happen with music.
Rob
So then is a simple test like feeding a sine wave in HF region and looking at hi x-o output through a scope not valid or useful?
Well, I'm gonna have to set it up again and retest with dummy loads and sinewave input. I believe I understand where Rob and I disagree, and that relates to whether the discharge path is the 2.2 meg or the source and/or load impedance. In the meantime, there's this, which may be read as saying Greg Timbers agrees with Rob:So then is a simple test like feeding a sine wave in HF region and looking at hi x-o output through a scope not valid or useful?
http://www.audioheritage.org/vbulletin/showthread.php?p=230425#post230425
d-ray657
Ain't this boogie a mess
Just sticking a little toe in to test the water on this whole e'wave thing. I have a few questions:
Is there a minimum size that will work for the project?
Can you achieve improvements with drivers over a broad range of quality - in other words, is it not worth the effort without something like the JBL drivers I have seen discussed in several posts?
I have a couple of very substantial cabinets that housed Fairfax 300x speakers. I replaced the tweeters when I first got them, and when I stupidly blew a woofer testing a bad cassette deck, I replaced the woofers with 10" Dynacos. The outer dimensions of the cabinets are 22" x 14" x 11." Would cabinets that small work with the e'wave concept? Would the Dynaco drivers suffice to provide the bottom half of the project? Finally, because of the relatively low profile of the cabinets, would placement on a short, slightly angled stand interfere with the direction of the sound waves?
Thanks for any information.
Regards,
D-Ray
Is there a minimum size that will work for the project?
Can you achieve improvements with drivers over a broad range of quality - in other words, is it not worth the effort without something like the JBL drivers I have seen discussed in several posts?
I have a couple of very substantial cabinets that housed Fairfax 300x speakers. I replaced the tweeters when I first got them, and when I stupidly blew a woofer testing a bad cassette deck, I replaced the woofers with 10" Dynacos. The outer dimensions of the cabinets are 22" x 14" x 11." Would cabinets that small work with the e'wave concept? Would the Dynaco drivers suffice to provide the bottom half of the project? Finally, because of the relatively low profile of the cabinets, would placement on a short, slightly angled stand interfere with the direction of the sound waves?
Thanks for any information.
Regards,
D-Ray
Welcome to the E'Wave adventure!
Nope, we have "official" E'Waveguides as small as 6" square now, and there's nothing inherent in the principles practiced here which would preclude E'Waving Minimus 7, even.
Understand that E'Wave is essentially a high-frequency section, a tweeter replacement, if you will, thus far primarily used in two-way systems. As tweeters go, it's a fairly expensive substitute at $168 nominal per pair, but it is much more than just that, rather, a high-performance engineered system incorporating features transcending what is possible with conventional designs which can transform existing systems. Several of us have been astonished by what occurs when E'Wave is mated with Advent woofers and inexpensive Daytons, for example; it'll work with any woofer that can play competently up into its nominal 1.6 kHz crossover region. That's not to say that the sonic character is not substantially dependent upon the quality of the low frequency portion, but many builders have been pleasantly surprised by the achievable results with what might otherwise be considered "pedestrian" woofers.
Absolutely.
Very likely. There have been discussions in this thread regarding designing with A-25 woofers, and one member has tried E'Wave with A-25 itself here, as well, to good outcome.
The vertical beamwidth of the standard waveguide is 50° with constant directivity, effectively wider than many tweeters. With the 6" square waveguide, this expands to 90°, which few other designs can even approach. If it's not possible or desireable to elevate the waveguide axis to listening height, aiming it there will easily suffice, yes.
You are very welcome, of course, and I hope you find this reply responsive to your concerns.... :thmbsp: