I told you - Wiki!!
Appears my sense of humour does not translate across the pond!
I told you - Wiki!!
Appears my sense of humour does not translate across the pond!
Sorry, my bad.
We were discussing, as we basically see eye to eye.
As to music, think of the speaker this way:
Push 8 volts peak into an 8 ohm nominal 3 way speaker at 30 hz. One ampere peak, 8 watt peak.
Now, push 8 volts at 1 kHz into same by itself. Again, 1 amp peak 8 w peak.
Now, 5khz 8 volts. 1 amp. 8 w peak.
Now, push all three signals at the same time. First, the amp needs 24 volts peak to push the three signals summed yet we only wanted 24 watts peak, the amp has to be capable of 3 amps, or 72 watts peak.
And if the woofer is reactive, you don't know where it will need it's peak current or peak voltage. There will be times when the speaker is full positive voltage but full negative current. That is actually the definition of reactance, as that is when the load is returning energy to the amplifier.
Sines are very predictable. Music is not supposed to be. How the pass devices handle the not so correlated output terminal voltage and output terminal current when three reactive loads are connected in parallel to the output node.... That is what that EDPR test is trying to relate to the customer.
Jn
I'm sorry, but the two things are perfectly correlated since the load is a linear system. A nonlinear system would still be mostly correlated with a small error.
Agreed - it did not seem clear to me. This is all classic linear systems theory, nothing new.
I told you - Wiki!!
Appears my sense of humour does not translate across the pond!
I recall when wiki wiki wiki was the name for a user-editable website, then it was shortened to just wiki, then there was a new user-editable encyclopedia named Wikipedia, then it became popular and people started calling it Wiki for short.
There are still wikis in the older sense, in fact if you look at the blue line near the top of this page, you will see ...
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki
Pun noted.
Someone should make a point of expunging all Wiki of simply wrong stuff. My wife showed me something funny, apparently some IT professionals spend a little spare time hacking on line scammers and make Youtube videos of it. Some are hilarious, even to the point of taking over their computers and refunding the scammed customers.
I am speaking about music. At any instant in time, the output node voltage and current will not be correlated. The best anyone can do is predict the I/V envelope area.
If you choose to describe a ten minute musical passage created by humans mathematically, be my guest. The computing horsepower required to exactly define the output node location in I/V space for all instances in time would be quite daunting, but could be calculated.as they are linear to a large degree.heck, I've done that in excel for three sines to examine the pass element dissipations, not difficult.
As it is, you only worry about single sine response and the simple calculations that entails.
Since you don't design the electronics, you certainly do not consider how your two or three way system exercises the output I/V space. And quite honestly, you shouldn't care. You have the luxury of ignoring what happens to the output devices in time.
The test described is an attempt to look into that concern, as people like you choose not to.
Again, your choice to ignore what your load can do to an amp is not a bad one in this day, as most amps are sufficiently designed to allow that choice. You simply choose to concentrate on a specific area and let others cover the amps. No problem.
Ps. As one with lots of experience pushing bipolars to the extremes of their SOA from DC to microseconds after bonding the chips to substrates and the wire bonds to the chip top, I have seen and tested all I speak of, so am quite aware of what the output node conditions under complex loads can do to chip dissipation. And I have lots of experience performing autopsies on said chips as well. That is where I learned all that base and emitter footprint stuff. Metallurgical stereoscopes are your friend..
I believe the test described is a reasonable one.
Jn
Last edited:
To see what I speak of, all one needs to do is connect an x/y scope to the output of an amp driving a 3 way with music. Horizontal voltage, vertical current.
If you drive a resistor, the display will be a line through 0,0, and pointing into both quadrant 1 and 3. In Q1, the positive pass elements will dissipate along a straight load line, in Q3 the negative elements dissipate.
If you put a reactance as load, you will see an ellipse or circle. This represents energy going to the load in Q1,3, and returning to the amp in Q2 and 4. In Q2 and 4, the pass elements see voltages higher than their own rails caused by the reactances, and it is there that the load returning energy is dissipated.
When you put a 3 way in place and play music, what you will see appears chaotic. There is a rhyme and reason to the display, but the designer and the user cannot worry about the exact location of the trace, but only the area the trace can take in the display.
The EDPR test is only a rudimentary attempt as it only considers one frequency at a time, but it is a start down a long learning curve.
Jn
If you drive a resistor, the display will be a line through 0,0, and pointing into both quadrant 1 and 3. In Q1, the positive pass elements will dissipate along a straight load line, in Q3 the negative elements dissipate.
If you put a reactance as load, you will see an ellipse or circle. This represents energy going to the load in Q1,3, and returning to the amp in Q2 and 4. In Q2 and 4, the pass elements see voltages higher than their own rails caused by the reactances, and it is there that the load returning energy is dissipated.
When you put a 3 way in place and play music, what you will see appears chaotic. There is a rhyme and reason to the display, but the designer and the user cannot worry about the exact location of the trace, but only the area the trace can take in the display.
The EDPR test is only a rudimentary attempt as it only considers one frequency at a time, but it is a start down a long learning curve.
Jn
I note your info still says Belmont mass. What up wit that? Edit:just changed mine.
Sigh, have to go to Colorado soon. Not looking forward to that.
Jn
Last edited:
I note your info still says Belmont mass.
Still???? That is where we ended up after the experience from hell. Well not quite, I admit I had several temporary addresses but this one is hopefully permanent (for a while at least). Funny when you settle out of court both lawyers take some of your money and both tell you that you won.
Again, not really true. One cannot talk about correlation at an instant in time. Correlations occur across time.
An I?V envelope is of use, I've used it myself, but it cannot show correlations.
Then you must not believe in superposition or the Fourier Transform, because they tell us that any system response can be viewed as a sum of sine waves. That's what convolution is, a summing of the individual sine waves to yield the result.
A simplistic response used in an attempt to divert from reality.
Reality is a driver modifying it's inductance as a function of position.
Reality is a driver transfer function being modulated, the output being a product of two frequencies.
Reality is an inductor introducing a third harmonic because of proximity induced resistive modulation and slew dependent eddy resistive losses at second harmonic.
In my line of work, I cannot afford the luxury of simplistic beliefs or assumptions such as yours. I learned Fourier, superposition, all you speak of long ago. I am trying to teach you what I deal with daily.
Yet you try to play word games. I thought more of you than that. Perhaps I was incorrect in my thinking.
As to the op, a 2 or 3 way can load the amp below the nominal impedance.
Jn
Reality is a driver modifying it's inductance as a function of position.
Reality is a driver transfer function being modulated, the output being a product of two frequencies.
Reality is an inductor introducing a third harmonic because of proximity induced resistive modulation and slew dependent eddy resistive losses at second harmonic.
In my line of work, I cannot afford the luxury of simplistic beliefs or assumptions such as yours. I learned Fourier, superposition, all you speak of long ago. I am trying to teach you what I deal with daily.
Yet you try to play word games. I thought more of you than that. Perhaps I was incorrect in my thinking.
As to the op, a 2 or 3 way can load the amp below the nominal impedance.
Jn
Earl,
My apologies.
I did not react well to you treating me as a troll.
My day job involves "stuff" decades beyond SOTA, so being questioned as to my understandings of simple linear theory, offended me. Edit: btw, I do not mean SOTA for audio, but for far more complex things.
For what I do, I am here to explain. For what I do not, I am here to learn.
At some point, you may decide that my understandings are of value and you will benefit. If not, that is not my concern.
I post for others to learn. If that happens, great. If not, I cannot worry. Life is too short.
John
My apologies.
I did not react well to you treating me as a troll.
My day job involves "stuff" decades beyond SOTA, so being questioned as to my understandings of simple linear theory, offended me. Edit: btw, I do not mean SOTA for audio, but for far more complex things.
For what I do, I am here to explain. For what I do not, I am here to learn.
At some point, you may decide that my understandings are of value and you will benefit. If not, that is not my concern.
I post for others to learn. If that happens, great. If not, I cannot worry. Life is too short.
John
What is lost is this...
Music is not steady state. It is transient by nature.
Bass reflex, not steady state. While swept sine says efficiency is this.... It is not true.
Group delay for multi order systems I do not care for. I prefer FL horns or critically damped sealed. Relying on a buildup of multi order, not crisp..4th, 6th order...great for long boomy bass notes, but not my cup of tea.
A speaker cab which is dependent on the buildup of energy is calculable w/r to response, but music can be (should be) unpredictable. Otherwise, just replace the humans with machines.
The I/V trace of music is not predictable, and for speakers with multi order response, would require multiple cycles to settle down to a linear predicted response.
For the OP question, simple linear is not enough. Developing an IV space for a specific system is needed if the loading can hit the foldback limits on modern amps.
Describing any speaker system using a swept sine, while useful, is not far enough. Eventually, some will go further.
Jn
Music is not steady state. It is transient by nature.
Bass reflex, not steady state. While swept sine says efficiency is this.... It is not true.
Group delay for multi order systems I do not care for. I prefer FL horns or critically damped sealed. Relying on a buildup of multi order, not crisp..4th, 6th order...great for long boomy bass notes, but not my cup of tea.
A speaker cab which is dependent on the buildup of energy is calculable w/r to response, but music can be (should be) unpredictable. Otherwise, just replace the humans with machines.
The I/V trace of music is not predictable, and for speakers with multi order response, would require multiple cycles to settle down to a linear predicted response.
For the OP question, simple linear is not enough. Developing an IV space for a specific system is needed if the loading can hit the foldback limits on modern amps.
Describing any speaker system using a swept sine, while useful, is not far enough. Eventually, some will go further.
Jn
Now THAT'S what I'm talking about...or could be. I gotta get some equipment together to test things. Though it seems even from a static curve, there are rough equivalent impedances presented to amplifiers. Maybe we need to back to the days when KEF showed a smiling Richard Small telling us they had made our amps twice as powerful!
When I mentioned that to him once he looked embarrassed, though considering some thread contents here he needn't have been methinks.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.