Amplifier Output Impedance (Damping Factor) and Speakers

[mhardy6647]

The advantage of these T+A amps is that only the output impedance can be changed and nothing else. I can't imagine a better UUT for this topic.

Variable output impedance was a "thing" a long, long time ago in hifi, for those who weren't aware of it.

source: https://www.alliedcatalogs.com/flipbook/1959_allied_radio_catalog.html
UL listed, too!

 

[fpitas]

Variable output impedance was a "thing" a long, long time ago in hifi, for those who weren't aware of it.
View attachment 330872
source: https://www.alliedcatalogs.com/flipbook/1959_allied_radio_catalog.html
UL listed, too!

Output impedance of Large, Really Large, and "OMG the cone will never stop ringing"

 

[sergeauckland]

Pye amongst others made amplifiers with adjustable positive feedback that could reduce the output impedance to zero, or even negative if one wasn't careful.

I incorporated those ideas into a 100 wpc amplifier I designed in the 1970s and it worked fine, best set to a small value (like 0.1 ohms) positive. Still not as low as SS amps, but a lot better than the amp without it.

S.

 

[Deleted member 48726]

Output impedance of Large, Really Large, and "OMG the cone will never stop ringing"

The Forever Bass setting..

 

[egellings]

So if I read an amp manufacturer says they increased DF from 240 to 320 between version 1 and version 2 of an amp, does that really matter?

Nope, unless other more important audibly noticeable characteristics got improved as a byproduct.

 

[MattHooper]

This repeats work I did long ago, and no doubt done by many, many others. It helps explains my position on amps, which is that for most speakers I find it easy to distinguish a SS from a tube amp, find it almost impossible to pick from two SS amps at reasonable levels, and find it much harder to pick between two tube amps (though IME tube amps tend to vary much more than SS amps). This also helps explain why some people prefer certain amp/speaker pairings.

Reviving this thread just for this question:

Don (or others)..,

If we keep this in the realm of those tube amplifier and speaker pairings where the tube amplifier actually does audibly modify the sound…

If we presume that the main thing people would be hearing are variations in frequency response when a tube amp is interacting with the impedance of a loudspeaker, this should be reproducible within an appropriate equalizer, right?

How many people seem to really like what a tube amplifier does when they swap it instead of a solid-state amp, and if it’s mostly just doing some modification to the frequency response, could it be the case that such two amplifiers tend to modify the frequency response in a sort of graceful way?

I mean, if one inserted an eq, random movements, up-and-down of narrow frequency, spectrums might not be likely to sound good.

So I’m asking if perhaps two amplifiers have a tendency to “eq” sound in a pleasing manner - perhaps with a certain Q factor behaviour that sounds smooth and pleasing to the ear?

I’m thinking, for instance, also of my experience hearing some modest Totem speakers at my friends place both on a Hegel SS amp and a 300B Fezz Audio tube amp.
I posted the measurements of that amplifier on another thread where people said that the amplifier is highly likely to modify the sound of a speaker depending on its impedance.
I found the sound so much more seductive with the 300B amp, smoother and more natural tone (to my ears). So at least me to wondering if there’s a certain manner in which Tube amps tend to modify frequency response that tense not to sound jagged.

(I asked that also knowing that there can be tube amp interactions with the speaker that lead to lean or bright sound as well)

 

[solderdude]

If we presume that the main thing people would be hearing are variations in frequency response when a tube amp is interacting with the impedance of a loudspeaker, this should be reproducible within an appropriate equalizer, right?

Yes but the output impedance of such amps may not be constant over the entire frequency range so would have to be characterized.
Then you need to know the impedance plot of the used speaker and combine that with the amp to arrive at an elaborate EQ and when applying room correction you would have to modify the EQ again. And then you might have audible amounts of distortion (Harmonic and IM as they go hand in hand)

could it be the case that such two amplifiers tend to modify the frequency response in a sort of graceful way?

Sure... usually the bass is elevated and the midrange is modified a little and maybe the highs as well.

 

[KSTR]

If we presume that the main thing people would be hearing are variations in frequency response when a tube amp is interacting with the impedance of a loudspeaker, this should be reproducible within an appropriate equalizer, right?

Indeed frequency response deviations can quickly be, and often will be, the dominant factor of perceived sound difference between solid-state and tube power amplifiers.

But once you have the exact same signal response at the speaker terminals two amplifiers will sound the same to at least first order, and that sameness can be dialed in with proper EQ (by measuring the impulse response of both amps at the speaker, then some math wizardry to arrive at a sophisticated fine-detailed EQ -- in practise a impulse response used as a convolution kernel9.

But the fine-print of the perceived sound may be different. You will have different distortion profiles vs. signal levels, distortions of all kinds.

 

[DonH56]

Reviving this thread just for this question:

Don (or others)..,

If we keep this in the realm of those tube amplifier and speaker pairings where the tube amplifier actually does audibly modify the sound…

If we presume that the main thing people would be hearing are variations in frequency response when a tube amp is interacting with the impedance of a loudspeaker, this should be reproducible within an appropriate equalizer, right?

Perhaps, but it may be more difficult than just adjusting the amplitude because phase also changes, both the amplifier and speaker/crossover system. "Appropriate" EQ could be very complex...

Note time domain response is also affected, including things like settling tails and envelope (attack, decay, sustain, release) response, that steady-state frequency analysis shown here does not show. So again it may not be so simple.

How many people seem to really like what a tube amplifier does when they swap it instead of a solid-state amp, and if it’s mostly just doing some modification to the frequency response, could it be the case that such two amplifiers tend to modify the frequency response in a sort of graceful way?

There are many other factors at play, including distortion characteristics, noise floor, and such that may cause people to prefer a particular sound. Tube or solid state, if we just address frequency response, people may prefer a particular amplifier because it produces a more desirable response coupled to a particular speaker. Interaction, synergy, whatever you choose to call it.

Distortion characteristics are often cited since many tube amplifiers have single-ended stages that increase even-order distortion, leading to second and low-order even harmonics dominating, compared to SS amplifiers that tend to have differential circuits that cancel even-order harmonics so odds dominate. SS amps usually have greater feedback and thus much lower distortion, but I have read that some people may actually prefer small amounts of low-order even distortion, leading to a preference for tube amplifiers. That is a different topic than this thread, however.

I mean, if one inserted an eq, random movements, up-and-down of narrow frequency, spectrums might not be likely to sound good.

Of course not, but the amplifier/speaker interaction is not random, so people can choose what they like to hear. Also note narrow frequency deviations can be much harder to detect than broad peaks or valleys.

So I’m asking if perhaps two amplifiers have a tendency to “eq” sound in a pleasing manner - perhaps with a certain Q factor behaviour that sounds smooth and pleasing to the ear?

I am not sure how you are relating Q factor in this context. I am also guessing you mean to say "tube" where "two" is written (auto-incorrect?) Speaking for myself, I find tube amplifiers sound good with some speakers, but poor on others, largely a function of the speakers' impedance characteristics. The curves I have shown for different speakers indicate how the response may be changed more significantly for speakers with a wider range of impedance values over frequency. That may sound better or worse to you (or anyone, a general "you"). Note EQ will not change the fundamental distortion characteristics of the amps, though of course can affect the amplitude of the harmonics (and fundamentals).

I’m thinking, for instance, also of my experience hearing some modest Totem speakers at my friends place both on a Hegel SS amp and a 300B Fezz Audio tube amp.
I posted the measurements of that amplifier on another thread where people said that the amplifier is highly likely to modify the sound of a speaker depending on its impedance.
I found the sound so much more seductive with the 300B amp, smoother and more natural tone (to my ears). So at least me to wondering if there’s a certain manner in which Tube amps tend to modify frequency response that tense not to sound jagged.

The opposite is more likely to be true; a tube amplifier's higher output impedance means the frequency response depends more upon the speaker, so large changes in impedance are more noticeable. This may do things like increase the bass and/or roll off the highs in a pleasing manner whereas a SS amp will provide a flatter frequency response.

(I asked that also knowing that there can be tube amp interactions with the speaker that lead to lean or bright sound as well)

I would argue that is a primary factor, as mentioned above, since a tube amp is more likely to be affected by the speaker's impedance curve than a SS amp. For better or worse.

HTH - Don

 

[SIY]

Perhaps, but it may be more difficult than just adjusting the amplitude

Or... you can do it the lazy way and just stick a resistor in series with the speaker. 99.9% sure that's all Carver did in his Stereophile stunt.

Yes, yes, this assumes that output impedance of a tube amp is constant with frequency. But that's very often the case, or nearly so, since feedback factors (and thus compensation) are much lower in most tube designs. And if there's a rise in impedance at the top end, that has less of an EQ effect with most dynamic speakers because of their rise in impedance in the same region. The resistor will have the greatest audible effect in the low end (because of the wildly swinging speaker impedance) and near crossover frequencies- and those are in the lower frequency areas where flat output impedance is a good approximation.

 

[DonH56]

Or... you can do it the lazy way and just stick a resistor in series with the speaker. 99.9% sure that's all Carver did in his Stereophile stunt.

Yes, yes, this assumes that output impedance of a tube amp is constant with frequency. But that's very often the case, or nearly so, since feedback factors (and thus compensation) are much lower in most tube designs. And if there's a rise in impedance at the top end, that has less of an EQ effect with most dynamic speakers because of their rise in impedance in the same region. The resistor will have the greatest audible effect in the low end (because of the wildly swinging speaker impedance) and near crossover frequencies- and those are in the lower frequency areas where flat output impedance is a good approximation.

It was way too late when I posted and I failed to mention this as the most obvious thing to try if you just want to make your SS amp sound similar to a tube amp. IME a series resistor works much better than trying to EQ the signal, though does not change the distortion or saturation characteristics of the amps (which is a factor if the tube amp has significantly lower power as is often the case). But again distortion is not the focus of this thread and way outside its scope.

 

[wwenze]