The discussion depends strongly on system components. Low source impedance would be more consequential with full-range speakers or active crossovers than typical multi driver commercial offerings. The iron core woofer inductors salvaged from PSB and Paradigm 2-ways recycled at work approached 0.5 ohm DCR. Ignoring speaker cables and connector losses even miracle amps are limited to a DF under 20 with these speakers.
So no one's mentioned deliberately adding resistance to make a SS amp more closely resemble a SE tube DF wise. Heck, it's only a few watts, why the "resistance" to the idea?
So where exactly does the amplifier output impedance "crossover" into current source behavior anyway, relative to nominal speaker impedance? 0<DF<1; I source, 1<DF<100; V source?
So where exactly does the amplifier output impedance "crossover" into current source behavior anyway, relative to nominal speaker impedance? 0<DF<1; I source, 1<DF<100; V source?
I agree totally. And actually, I think the situation is even worse than is described in that article by Augspurger that you linked to.
As he correctly points out, the DC resistance of the loudspeaker should be added to the output impedance of the amplifier, and thus it belongs in the denominator in the calculation of "loudspeaker impedance divided by source impedance" when calculating the true damping factor. But Augspurger is making the mistake of still including it also in the numerator, since he feeds in 8 ohms for the loudspeaker impedance in his "Over-All DF" column in his Table 1.
The proper discussion gets rather tediously complicated because of the fact that the loudspeaker (any maybe also the amplifier) does not have a purely resistive impedance. Just to give the flavour of the point I am wanting to make, let's just pretend for now that everything is purely resistive. So, making this approximation, we have
Amplifier: Output impedance = Ra
Loudspeaker: Impedance = Z + Rs
Here, Rs is the DC resistance of the speaker coil, and Z is the intrinsic impedance of the speaker (the impedance it would have if the voice coil were made of superconducting wire).
The usual "naive" damping factor is calculated as
Dn = (Z+Rs)/Ra
The "Over-All DF" Augspurger is defining is
Do = (Z+Rs)/(Ra+Rs)
But the true damping factor should actually be
Dt = Z/(Ra+Rs)
In other words, Augspurger correctly adds the speaker resistance Rs into the denominator when calculating the damping factor. But he still also leaves it in the numerator as well, and it doesn't belong there at all.
Again, I want to emphasise that I have made substantial over-simplifications and approximations here by pretending that the intrinsic speaker impedance just combines additively with the speaker coil resistance, and so on. But the essential point I want to make is that in a correct damping factor calculation, the speaker resistance should be removed from the numerator, as well as being added in to the denominator. This means that the "true damping factor" is actually a lot smaller than even Augspurger is saying.
Well said. A loudspeaker & an amplifier are a system.
Now if you look at things statistically, low Rout amplifiers & speakers designed for then with little or no care to the end loudspeaker load dominate the current marketplace. It is easier, most of those amps don’t much care about the whoops in the impedance response.
Now, as Rout rises we see the bumps in the impedance response start to show up in the FR. Sometimes this effect will improve things, most often not. Some speakers are actually designed to be driven by a high Rout device.
Further, and this to the point made about improvements in mid/top, up where the Rout only changes FR above where the bass driver/box may (or may not) need “control", there is some buzz that current amps can sound better. This can lead to mixed amps in a multiamp situation, where a voltage amp is used at teh bottom where the “control” is needed and a higher Rout amplifier(s) up top where, without passive XO, the driver impedance can be essentially flat.
As far as bass goes, each speaker will have an optimum Rout amplifier, in most cases, but certainly not all, the Rout is low.
One has to keep an open mind.
As to analog_sa’s assertion that in his system, 20/30/50 damping factor is significant, i pulled the number 20 because that is what some luminary suggested in a technical paper. One has to make a call, the actual number will depend on the speaker’s Z (including the often seen big inductor in series with teh woofer), the room, room placement, listener taste, and the rest of the kit in the system.
dave
That would be a nominal 4Ω speaker. Having measured teh DCR of thousands of drivers nomibal 8Ω drivers typically measure 6.5-7.5Ω.
dave
Yes, we did not like what it did to the rest of the range (bass was improved). We decided after that, best to just use the right driver/box/amplifier.
dave
jjasniew,
That is a Great Idea!
Over 50 years ago, I did use a series resistor with a solid state amplifier speaker output terminal.
I had a Dyna kit SCA80 solid state amplifier, and a pair of KLH33 loudspeakers (10 Inch; lightly acoustic-resistive loaded port; and a phenolic tweeter).
I also mentioned the idea of using a series resistor with a solid state amplifier, more than once on these Tubes / Valves threads, because the subject of damping keeps coming up; especially in regard to tubes versus solid state.
The subject never dampens down, it just keeps going on (pun intended).
I recommend if you have a solid state amplifier, that you at least try it with several different resistance values.
There is nothing against real world tests . . .
“In theory, there is no difference between theory and practice.
But, in practice, there is”
And, a diversion/misquote from George Orwells novel "Animal Farm" . . .
"All loudspeakers are created equal, but some are more equal than others".
That is a Great Idea!
Over 50 years ago, I did use a series resistor with a solid state amplifier speaker output terminal.
I had a Dyna kit SCA80 solid state amplifier, and a pair of KLH33 loudspeakers (10 Inch; lightly acoustic-resistive loaded port; and a phenolic tweeter).
I also mentioned the idea of using a series resistor with a solid state amplifier, more than once on these Tubes / Valves threads, because the subject of damping keeps coming up; especially in regard to tubes versus solid state.
The subject never dampens down, it just keeps going on (pun intended).
I recommend if you have a solid state amplifier, that you at least try it with several different resistance values.
There is nothing against real world tests . . .
“In theory, there is no difference between theory and practice.
But, in practice, there is”
And, a diversion/misquote from George Orwells novel "Animal Farm" . . .
"All loudspeakers are created equal, but some are more equal than others".
Last edited:
A reasonably high DF lets the amp do a better job of putting the brakes on the cone when the signal has ended. If you have a raw driver lying around, thump the cone with the driver's input terminals open circuit and listen, and then thump the cone with a dead short across the input terminals. That represents a very exaggerated DF, more than you'll see at the amplifier's output terminals, but you'll get a good impression of damping's effect.
Combine a Hi-Q series 2 element low pass crossover filter, or a Hi-Q series 2 element high pass filter, with a very high damping factor amplifier.
Now, the driver impedance is the only damping on that Hi-Q circuit.
The driver impedance IS the Q determining factor.
If the total system Q is high . . . Ring!
Now, the driver impedance is the only damping on that Hi-Q circuit.
The driver impedance IS the Q determining factor.
If the total system Q is high . . . Ring!
But it can be too much.
Box. Driver.
A higher Qms driver will typically have a larger impedance peak at resonance and will rely more on electrical damping to control the damping.
Not always, the Fostex FE-series drivers like to see. high Rout amplifier, control being exerted by the (really required) horn loading it.
A driver with a low Qms will not care as much, but can become overdamped more easily.
The impedance bumps up higher in frequency (usually due to an XO) that annoy a high Rout amplifier, are a different story. Same problem, different cause.
dave
The whole discussion of DF is a sidetrack of reality.
What gets the cone moving is amps.
See Current-Drive - The Natural Way of Loudspeaker Operation
This makes the whole discussion of DF a pseudo discussion that lacking physic substance.
What gets the cone moving is amps.
See Current-Drive - The Natural Way of Loudspeaker Operation
This makes the whole discussion of DF a pseudo discussion that lacking physic substance.
Heh;
that's why for real damping my SE amps have a possibility to have a negative output resistance.
Yes, DF number is a hoax.
Here is the video that shows a negative output resistance: Wavebourn - Structure of Magic | Facebook
that's why for real damping my SE amps have a possibility to have a negative output resistance.
Yes, DF number is a hoax.
Here is the video that shows a negative output resistance: Wavebourn - Structure of Magic | Facebook
Last edited:
I think a reasonable mount of DF is necessary to control woofer motion. My amplifier has a DF of about 25, and that seems to work well. It may be that some speakers work better with differing amounts.
"There are many ways to get to Rome" 🙂
Many tube amps may have poor response at very low frequencies, but my tube amp offers a Freq. Response (+/- 0.5dB) 10Hz-30kHz, but they are "only" 38 watts, so I take the signal from the OPT output, process it below 40 Hertz with DSP, amplify with Class D, (1000 Watts bridge or 2x 450 watts stereo) and reproduce with 2 x 18 "subwoofers 150 liters each.
I'm very satisfied with the result.
DiaLogue Two — PrimaLuna USA
Damping Factor is just a (bad) inversion of Rout into a big number for the marketing guys to put on a spec sheet. The bigger the better. Just like the lower the THD is the better. We all (should) know that THD is really a fairly meaingless number.
dave
dave
I hear this alot. One can try (line level) equalising the response variations and see that nothing has fundamentally changed with the speaker. Rather than the speaker losing control, there simple interaction between Zo and speaker impedance, and the speaker response equates to the electrical response applied to it.
A lot of the amps with ridiculously high damping factors are also really, really big... so that may affect what gets heard more than anything else. People often rave about the low frequency performance on the BSS EPC780, and inevitably they claim it's because the advertised damping factor is greater than 1000.
Reality? It's probably that it's an extremely powerful amplifier (unlikely to clip and is capable of delivering an insane amount of current), with a Sowter input transformer.
In many cases, I think that a lot of people blame the bass performance of SE amplifiers on the damping factor, when in fact it may be due to excessive distortion as the amp starts to reach power levels where it is less linear.
Reality? It's probably that it's an extremely powerful amplifier (unlikely to clip and is capable of delivering an insane amount of current), with a Sowter input transformer.
In many cases, I think that a lot of people blame the bass performance of SE amplifiers on the damping factor, when in fact it may be due to excessive distortion as the amp starts to reach power levels where it is less linear.
I agree. The Krell KSA250 was perhaps the amp with the most impressive bass i've had in my system. It was obvious at any loudness level. Apparently the massively parallel output and the 4.5kVA transformer played a role yet damping factor was a very modest 60 from my measurements.