I know a few things about damping factor but not enough... I know it makes your bass sound flabby, booooomy, slow, like it's getting in the way of the music that's coming because the damn woofer is still recovering. I know physically, bad damping is the amps inability to control the momentum of the speaker cone, to slow it when it needs slowing, right? So it's kind of like the brakes on a car?
Is bad damping factor caused because tubes are inherently easier to "turn on" than they are to "turn off" (non linear)? If tubes are slower to turn off than turn on I could see how that might make it better at making the speaker cone move than making the speaker cone slow down. Is that right? Or am I wrong on that. Some other fundamental cause?
What is the best method to improve damping factor in an amp design? And, how far can/should you take that method before you start to cause some other problem in the design, because everything is a compromise right.
Is bad damping factor caused because tubes are inherently easier to "turn on" than they are to "turn off" (non linear)? If tubes are slower to turn off than turn on I could see how that might make it better at making the speaker cone move than making the speaker cone slow down. Is that right? Or am I wrong on that. Some other fundamental cause?
What is the best method to improve damping factor in an amp design? And, how far can/should you take that method before you start to cause some other problem in the design, because everything is a compromise right.
Speakers do the booming, and the amp can reduce that by effectively shorting the speaker terminals with
low source impedance. The DF is a small signal characteristic, and has little to do with output stage nonlinearity.
The DF number depends on the speaker impedance, so Zout is a better figure of merit for the amplifier.
Most amplifiers increase the DF (reduce the internal impedance) with high negative feedback.
Reducing the internal impedance with more paralleled output devices is more difficult for the driver stage.
Increasing the open loop gain in order to increase the NFB reduces bandwidth and stability.
low source impedance. The DF is a small signal characteristic, and has little to do with output stage nonlinearity.
The DF number depends on the speaker impedance, so Zout is a better figure of merit for the amplifier.
Most amplifiers increase the DF (reduce the internal impedance) with high negative feedback.
Reducing the internal impedance with more paralleled output devices is more difficult for the driver stage.
Increasing the open loop gain in order to increase the NFB reduces bandwidth and stability.
Last edited:
First, damping is a poor way of specifying the output impedance of the amplifier, Rout. Damping = (speaker impedance)/Rout so when Rout goes up damping goes down.
Marketers like using damping because they can use big numbers.
Highish Rout DOES NOT "makes your bass sound flabby, booooomy, slow, like it's getting in the way of the music”. If that happens it means your loudspeakers are underdamped for use with the amplifier in question.
Amplifiers, loudspeakers (and the connecting cable used) are a system and have to be considered as such.
If an amplifier has a highish Rout you do have to be more careful with loudspeaker choice, since the amplifier no longer ignores wild deviations in impedance response of the loudspeaker.
For example: http://wodendesign.com/downloads/King-of-Swingers.pdf
dave
Marketers like using damping because they can use big numbers.
Highish Rout DOES NOT "makes your bass sound flabby, booooomy, slow, like it's getting in the way of the music”. If that happens it means your loudspeakers are underdamped for use with the amplifier in question.
Amplifiers, loudspeakers (and the connecting cable used) are a system and have to be considered as such.
If an amplifier has a highish Rout you do have to be more careful with loudspeaker choice, since the amplifier no longer ignores wild deviations in impedance response of the loudspeaker.
For example: http://wodendesign.com/downloads/King-of-Swingers.pdf
dave
More NFB will cure some of the "booming"
( increased feedback will decrease output impedance and reduce distortion. )
no, its more like the shock absorbers in your car;
their duty is to absorb the energy out of the spring/mass system, convert it to heat, and thereby stop any unwanted movement; without shocks your car would be swinging up and down and oscillate out of control ...
Damping factor goberns the rate of decay of any mechanical or electrical circuit when it is suddenly exited by a shock. For example, hitting a bell will produce a sound that more slowly or more quicly it tends to dissapear as the vibration of it is lost or attenuated because of air viscosity.
The same occurs if you go from a well illuminated place to an obscure site or vice versa. Our eyes, whose sentitivity is logarithmic, meeds an amount of time to return vission to normality.
If in a tank with quiet water you throw a stone, the waves formed at the surface dampes out also because of viscosity of water.
Strictly speaking, as the decay of a magnitude of those mentioned are negative exponentials, they never dissapear completelly, or take infinite amount of time to dissspear. In the equation for this behavior it is where damping factor rises.
The same occurs if you go from a well illuminated place to an obscure site or vice versa. Our eyes, whose sentitivity is logarithmic, meeds an amount of time to return vission to normality.
If in a tank with quiet water you throw a stone, the waves formed at the surface dampes out also because of viscosity of water.
Strictly speaking, as the decay of a magnitude of those mentioned are negative exponentials, they never dissapear completelly, or take infinite amount of time to dissspear. In the equation for this behavior it is where damping factor rises.
The term "damping factor" has a built-in misleading effect. The actual resistive damping comes from the sum of the amplifier's output "resistance" plus the speaker's (including wiring and crossover network) resistances. For example, a nominally 8 Ohm speaker with 6 Ohms DC resistance, measured with a DVM at the amplifier end of the speaker wire, and an amplifier with a "damping factor of 8 (meaning 1 Ohm output resistance) has a damping to the speaker's movement of 7 Ohms. The same speaker with an amplifier with "damping factor" of 80 (meaning 0.1 Ohm output resistance) has a damping to the speakers movement of 6.1 Ohms. Not much difference.
All good fortune,
Chris
All good fortune,
Chris
The amplifier output impedance not only affects the woofer, it can affect the complete frequency response of the speaker.
Suppose a perfect amplifier has zero Ohms output impedance.
Then, connect a 1 Ohm resistor to the "8" Ohm amplifier output tap, and connect the other end of the 1 Ohm resistor to the speaker "8" Ohm input (and connect amp common to speaker common).
That perfect amplifier plus the series 1 Ohm resistor, would give the amplifier a specified damping factor of 8 (referred to 8 Ohms).
Take an "8" Ohm speaker that has a wide varying input impedance, many speakers do.
6 Ohms at 20Hz; 25 Ohms at 50Hz, 6 Ohms at 200Hz, 16 Ohms at 1.5kHz, 8 Ohms at 5kHz, and 12 Ohms at 20kHz.
The 1 Ohm resistor in series with the speaker impedance forms a voltage divider:
Let the amplifier output 1V at these frequency tones, one at a time, from 20 to 20kHz.
1 and 6 The voltage at the speaker terminals is 6/7 V @ 20Hz
1 and 25 The voltage at the speaker terminals is 25/26 V @ 50Hz
1 and 6 The voltage at the speaker terminals is 6/7 V @ 200Hz
1 and 16 The voltage at the speaker terminals is 16/17 V @ 1.5kHz
1 and 8 The voltage at the speaker terminals is 8/9 V @ 5kHz
1 and 12 The voltage at the speaker terminals is 12/13 V @ 20kHz
So the minimum voltage at the speaker is 6/7 V (0.857V)
and the maximum voltage at the speaker is 25/26 V (0.9615V)
**** Not much difference in voltages versus frequency (about 0.999dB). ~ 1 dB
Now, use a 4 Ohm resistor instead of the 1 Ohm resistor.
The damping factor, referred to 8 Ohm, is 2.0
4 and 6 The voltage at the speaker is 6/10 V (0.600V)
4 and 25 The voltage at the speaker is 25/29 V (0.862V)
**** The difference in voltages versus frequency across the speaker terminals is worse (about 3.15 dB) ~ 3 dB
Suppose a perfect amplifier has zero Ohms output impedance.
Then, connect a 1 Ohm resistor to the "8" Ohm amplifier output tap, and connect the other end of the 1 Ohm resistor to the speaker "8" Ohm input (and connect amp common to speaker common).
That perfect amplifier plus the series 1 Ohm resistor, would give the amplifier a specified damping factor of 8 (referred to 8 Ohms).
Take an "8" Ohm speaker that has a wide varying input impedance, many speakers do.
6 Ohms at 20Hz; 25 Ohms at 50Hz, 6 Ohms at 200Hz, 16 Ohms at 1.5kHz, 8 Ohms at 5kHz, and 12 Ohms at 20kHz.
The 1 Ohm resistor in series with the speaker impedance forms a voltage divider:
Let the amplifier output 1V at these frequency tones, one at a time, from 20 to 20kHz.
1 and 6 The voltage at the speaker terminals is 6/7 V @ 20Hz
1 and 25 The voltage at the speaker terminals is 25/26 V @ 50Hz
1 and 6 The voltage at the speaker terminals is 6/7 V @ 200Hz
1 and 16 The voltage at the speaker terminals is 16/17 V @ 1.5kHz
1 and 8 The voltage at the speaker terminals is 8/9 V @ 5kHz
1 and 12 The voltage at the speaker terminals is 12/13 V @ 20kHz
So the minimum voltage at the speaker is 6/7 V (0.857V)
and the maximum voltage at the speaker is 25/26 V (0.9615V)
**** Not much difference in voltages versus frequency (about 0.999dB). ~ 1 dB
Now, use a 4 Ohm resistor instead of the 1 Ohm resistor.
The damping factor, referred to 8 Ohm, is 2.0
4 and 6 The voltage at the speaker is 6/10 V (0.600V)
4 and 25 The voltage at the speaker is 25/29 V (0.862V)
**** The difference in voltages versus frequency across the speaker terminals is worse (about 3.15 dB) ~ 3 dB
Last edited:
Thanx, something i neglected.
In short, if the impedance rises so does the output power so those frequencies are louder. The “boomy” bass thing is related to the typical resonance peak as a loudspeaker rolls off. But speakers with XOs often have rollercoaster impedance and you will hear it. Here is a commercial high sensitivity speaker. If driven with a highish Rout amplifier the rise at the bottom and th emountain in the midrange will be heard.
dave
planet10,
Thanks!
Nice low impedance dips that go below the 4 Ohm line . . . about 3.6 Ohms, 2.8 Ohms, and 3.8 Ohms.
So much for a speaker nominal rating of 8, 12, or 16 Ohms; whichever Klipsch model that was.
Thanks!
Nice low impedance dips that go below the 4 Ohm line . . . about 3.6 Ohms, 2.8 Ohms, and 3.8 Ohms.
So much for a speaker nominal rating of 8, 12, or 16 Ohms; whichever Klipsch model that was.
What if the amplifier "knew" the impedance curve of the speaker ahead of time, such that it could provide a mirror image output impedance to the speaker curve? Put the high damping where it's useful, relax it where it's not needed.
I personally wouldnt know how to design an amplifier with flat frequency response into 8 Ohms, while at the same time, its output impedance varied a couple of decades in value over frequency. Using DSP, one could imagine an amplifier that first measures the speaker impedance, then flips that curve to effect its output Z.
Maybe that's as far as you can take it - if one could go there at all.
Is there a standardized method of measuring amplifier damping factor or internal impedance?
Manufacturers seem to publish specs without conditions such as frequency, power level, or load Z, etc.
Manufacturers seem to publish specs without conditions such as frequency, power level, or load Z, etc.
Resistance of the speaker cable offsets the improvement of DF beyond some point. You get diminishing return with increasing the DF. In my opinion, DF about 50 with an 8ohm speaker and 100 with a 4ohm speakers is good enough. Not much difference above that.
The damping directly is very complex. ITOH you can easily messure amplifier's output impedance using the definition: leave the amplifier's input shorted. Attach at the load of it a variable low value rheostat. Inject a signal into the output terminals of the amp through this rheostat. Attach an oscilloscope across amplifier terminals and to the rheostat. Use any safe value of audio signal. Increase or decrease the rheo's value until the voltage drop across rheo and amp be equal in manitude. Disconnect all and measure rheo value: it be the same as amplifier's output impedance.
As mentioned before, DF is another way to declare the internal output resistance Ro (at LF) of the amplifier. I use this procedure to measure it. Connect 8ohm 10W resistor on the output, set a test signal with desired frequency on the input and adjust the amplitude to get 1-2W on the output. Write down the voltage measured on the resistor terminals (let's call it Vo8). Change the resistor with a 4ohm one but keep the input signal amplitude unchanged. Measure the voltage on the resistor terminals again. This is Vo4.
Amplifier's output resistance is:
Ro = 8 * (Vo8 - Vo4) / (2 * Vo4 - Vo8)
Accordingly, DF = 8 / Ro
Amplifier's output resistance is:
Ro = 8 * (Vo8 - Vo4) / (2 * Vo4 - Vo8)
Accordingly, DF = 8 / Ro
The late, great Hi-Fi World looked into damping factor here: https://hi-fiworld.co.uk/av-receive... most loudspeakers we find in listening tests.
Here's an example of what can happen with a highish amplifier source impedance. It isn't just important at LF.
https://www.stereophile.com/content/octave-jubilee-mono-se-monoblock-power-amplifier-measurements
https://www.stereophile.com/content/octave-jubilee-mono-se-monoblock-power-amplifier-measurements
Attachments
IMO, the amplifier's output impedance altering the speaker crossover behavior is the biggest reason why tube amplifiers sound different.
Ed
Ed
DF of only 4 in this example is indeed noticable acousically.
But DF 500 is hardly distinguishable from DF 80.