|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.122.72.31
In Reply to: RE: output impedance posted by hitsware on May 05, 2014 at 16:09:42
That is a common means that works fine for a high impedance source but in a power amp seems to be a flawed means. Wouldn't you want to see that it makes half Power rather than half voltage? That way the impedance of the circuit under test will be equal to the known impedance of the load; where half the power generated by the output section is dissipated in the load and the other half in the output section itself.
Here are the numbers from the two methodologies (the Voltage and Power paradigms).
For the M-60, the impedance of the output section is about 4 ohms, as predicted by the formula and confirmed by measurement. The Voltage method gets more like about 16-17 ohms.
The M-60 makes its maximum power into 16-17 ohms, which suggests that its actual impedance must be lower.
Follow Ups:
"For the M-60, the impedance of the output section is about 4 ohms, as predicted by the formula and confirmed by measurement. The Voltage method gets more like about 16-17 ohms."
Do you have a reference to the formula you are referring to, that gives 4 ohms? I'd be curious to see how that works.
Chriis
Gnd-----Volts-----OutputZ-----Measure-----LoadZ------GndIt seems obvious that Measure will be Volts/2 when
LoadZ = OutputZ
Also that the same Power will be dissapated in both Load
and Output
Perhaps my schooling was bunk ? .... :)
(unless there is a powerful reactive component somewheres)
Edits: 05/06/14
Your argument seems sound to me. It is reasonable to model the output of the amplifier as an ideal (zero-impedance) voltage source in series with a resistor R0. Here R0 is what one would call the "output impedance."
This should be a good approximation as long as the amplifier is not being asked to deliver more current than it is capable of supplying. (If more current than that is asked for, then the amplifier will be overloading and clipping horribly, and so discussions of output impedance then become essentially irrelevant and academic.) So in any measurement, it should always be ensured that the amplifier is not pushed beyond clipping.
There are various ways that R0 in the above model could be measured. One way would be as you said, to connect a load R and adjust R until the measured output voltage is half of the open-circuit output voltage. Another way would be to adjust R until the power dissipated in R is maximised (at fixed output level). The answer in either case will be the same, namely when R is equal to R0. (And the power dissipated in R will equal the power dissipated in R0, if R is set equal to R0.) And so whether judged by voltage or by power, the conclusion would be the same, that the output impedance is R0.
If the M-60 delivers maximum power into 16-17 ohms, then that seems to support the argument that the output impedance is 16-17 ohms.
Chris
Having designed speakers on Ralph's amps, my measurements are consistent with his claims for output impedance. For instance I can look at the bass response curves with a high-damping-factor solid state amp, and then with his amp, and see that the change is consistent with the Qes modification predicted by Ralph's output impedance claims.
I'm not EE enough to debate the theoretical validity of his technique, but I'm speaker designer enough to confirm its results via measurements.
Me being a dealer makes you leery?? It gets worse... I'm a manufacturer too.
The technique is quite solid when no current is involved.
I have to assume though that if there is a resistance that is used as a load that the voltage can only be there due to the fact that current is flowing. So power is involved.
Now if that is the case, you can't have power with no load, right? So you would start with some very high impedance well outside the circuit's range, and decrease the impedance until a peak in power is observed. Now we know the maximum power the circuit can produce. To sort out the output impedance, it will be equal to the load impedance when equal amounts of power are dissipated by the load as is by the output circuit. This point is of course when the load impedance equals that of the internal impedance that is dissipating the other half of the power.
This is how it was done in the old days. If you were educated in audio after the 1970s, its likely that the method you proposed would have been taught. You might want to look at this link to see what is going on.
The Voltage Paradigm has a number of 'charged terms' which don't quite mean the same thing as output impedance. I can give you a wonderful example of how this can affect your viewpoint.
If you were taught this method, you were probably also taught that applying loop negative feedback will lower the output impedance. But have you ever considered that such would violate Kirchoff's Law?
3.125 Watts = 25 Volts / 8 Ohm so peak power is where Zout = Zload
and Vload = Vnoload / 2
Edits: 05/06/14 05/06/14
Which is neither a good or bad thing.
The reason I use 'paradigm' is such exists as a body of thought, outside of which everything else appears incorrect or even blasphemous.
FWIW I did not create the Power Paradigm- it existed before I was born. Google 'Fisher A-55 amplifier' and look at the first hit, which is a YouTube image. The setting at 12 noon on the control is 'Constant Power', at fully CCW it is 'Constant Voltage'. At fully CW the control is marked 'constant current' but that never developed as a common method of driving loudspeakers.
Amplifiers on the voltage paradigm tend to have very low output impedances, those on the power paradigm tend to have output impedances in the range of or just below that of the loudspeaker impedance, constant current amplifiers tend to have output impedances that are some multiple of the impedance of the load.
This all has to do with how the speaker is damped; in the old days there was not a set formula for that; so matching speakers and amps was done differently. That is why older speakers (EV, Altec, JBL and other horns) usually have midrange and treble controls. They are not there to adjust the speaker to the room, they are there to adjust the speaker to accommodate the voltage response of the amp.
(Loudspeakers that operate with open baffle designs are best driven by amps that provide little or no damping and might require only 1:10 as a damping factor in the amp)
EV and MacIntosh led the way to create the idea of making the amplifier into a voltage source; in this way speaker response would be predictable with greater plug and play. This worked quite well with box speakers (although the Acoustic Research AR-1, the world's first acoustic suspension loudspeaker, was designed for an amplifier with a 7-ohm output impedance and was an example of a Power Paradigm device), although not so well with horns, ESLs and planar loudspeakers whose impedance curve does not derive from the resonance of a driver in a box.
However the voltage paradigm pretty well took over in the 1960s and 70s, with transistor amps being more common and with which it was easier to create a constant voltage response. Along with it our methodology of measurement changed- as in output impedance here, also we see the spec of speaker Sensitivity as opposed to Efficiency (which are the same thing only when the speaker is 8 ohms).
FWIW no loudspeaker needs a damping factor of over 20:1 and most are best damped with something less. A link to the article written by the head engineer of EV is below.
I guess so as calculation
BUT I don't subscribe to super
low Z out amplifiers. My main
circuit is ~ 1 Ohm out (channel
resistance of mosfet followers)
(no global feedback)
I'm working on a floating supply
circuit with ~ 80 Ohm out, (sans
feedback), but so far can't get
the hum out :( (I hate when that
happens)
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: