advertisement


Electrical Output Impedance (damping factor)... again!

These measurements allow us to compare the response of the amplifier into both a flat and a simulated speaker load (black trace) with 10dB feedback or with no feedback

https://www.stereophile.com/content/cary-audio-cad-805rs-monoblock-power-amplifier-measurements

119C805fig02.jpg

Fig.2 Cary CAD-805RS, 845 tube, 10dB feedback, frequency response from 4 ohm tap at 2.83V
into: simulated loudspeaker load (gray), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

119C805fig03.jpg

Fig.3 Cary CAD-805RS, 845 tube, 0dB feedback, frequency response from 16 ohm tap at 2.83V
into: simulated loudspeaker load (gray), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

Really bad design there with the peaking at LF when NFB used!
 
Most engineers who are not old enough to have grown up with valves would not know that LF instability exists.
It is very rare to see a reviewer checking output power at LF either. Most OP transformers are too small and saturate
 
Very generally speaking, a flat-impedance speaker or headphone would be expected to produce an "upside down U-shaped" frequency response when driven by an amplifier that has a "U-shaped" output impedance curve, compared to an amp with a flat impedance curve, is that correct?

Thus, if an amplifier with "U-shaped" output impedance curve is paired with a speaker or headphone that also has a "U-shaped" impedance curve, the deviations in output frequency response should cancel out to some degree (depending on how the two impedance curves align etc), is my logic correct?

So, if a valve amp has highest output impedance at low bass and high treble frequencies, it should be preferable to partner it with a speaker or headphone that also has high impedance at these frequency extremes, to reduce the extent of FR roll-off?
 
Thus, if an amplifier with "U-shaped" output impedance curve is paired with a speaker or headphone that also has a "U-shaped" impedance curve, the deviations in output frequency response should cancel out to some degree (depending on how the two impedance curves align etc), is my logic correct?
You will never manage to design a match. Some combinations get lucky​
 
Food for thought:

The Amplifier Damping Factor and its Application to Speaker Performance

https://www.pearl-hifi.com/06_Lit_Archive/02_PEARL_Arch/Vol_05/Sec_23/1385_Critical_LS_Damping.pdf

That is a very informative article, it's hard to believe it was written 65 years ago! I wonder how much of it is still relevant given how technology has moved on, e.g. more powerful loudspeaker magnets. Also, one would assume that most contemporary loudspeaker designs (i.e. excluding high-efficiency low-excursion designs), are voiced with low amplifier output impedances in mind, and so presumably the risk of overdamping is much lower?
 
Yes, I suspect the norm for speaker designers is to use an amp with a 'low' (compared with the speaker) output impedance when developing their speakers. Thus implicitly that is the behaviour they expect/assume and go for.

Measuring the impulse response at the speaker terminals is of limited used if you're listening to the speaker *output*. Reason being that the time domain of most speakers in most rooms is waaaay more severe and complicated than you'd seen at the terminals. i.e. The real problems are after those terminals.

Bottom line, yes, shoving a passive network between the amp and speaker can act as a 'tone control' and you may like the result when someone else dislikes it.

Personally, though, I always felt that having tone controls on the amp made more sense as the user can twiddle them to suite each item of music. The reality, again, being that these tend to vary as much as speakers and rooms. :)
 
Personally, though, I always felt that having tone controls on the amp made more sense as the user can twiddle them to suite each item of music. The reality, again, being that these tend to vary as much as speakers and rooms.

I’ve always felt it is something rather different than bass quantity, more a speed/decay/bloom sort of thing. As an example an ancient high-efficiency valve-era speaker such as my Klipsch La Scalas usually sound rather small, pinched, flat, 2d and dry with many solid state amps. They just sound stifled and don’t breathe. Even if one turns up the bass they still don’t do what they should, but put a period-appropriate little valve amp with them and they grow in stature, scale and naturalness. They just relax and flow. These speakers, like all others, were clearly designed with the equipment of their era and just don’t sound right with modern high-power amps designed to throw heavy plastic cones a very long way.

Given vintage speakers and high-efficiency horns are quite a big ‘thing’ at present in certain circles I suspect there might even be a market for an amp with a ‘damping factor’ control!
 
I’ve always felt it is something rather different than bass quantity, more a speed/decay/bloom sort of thing. As an example an ancient high-efficiency valve-era speaker such as my Klipsch La Scalas usually sound rather small, pinched, flat, 2d and dry with many solid state amps. They just sound stifled and don’t breathe. Even if one turns up the bass they still don’t do what they should, but put a period-appropriate little valve amp with them and they grow in stature, scale and naturalness. They just relax and flow. These speakers, like all others, were clearly designed with the equipment of their era and just don’t sound right with modern high-power amps designed to throw heavy plastic cones a very long way.

Given vintage speakers and high-efficiency horns are quite a big ‘thing’ at present in certain circles I suspect there might even be a market for an amp with a ‘damping factor’ control!

There's this one:

These measurements allow us to compare the response of the amplifier into both a flat and a simulated speaker load (black trace) with 10dB feedback or with no feedback

https://www.stereophile.com/content/cary-audio-cad-805rs-monoblock-power-amplifier-measurements

119C805fig02.jpg

Fig.2 Cary CAD-805RS, 845 tube, 10dB feedback, frequency response from 4 ohm tap at 2.83V
into: simulated loudspeaker load (gray), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

119C805fig03.jpg

Fig.3 Cary CAD-805RS, 845 tube, 0dB feedback, frequency response from 16 ohm tap at 2.83V
into: simulated loudspeaker load (gray), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).
 
I’ve always felt it is something rather different than bass quantity, more a speed/decay/bloom sort of thing.


Given vintage speakers and high-efficiency horns are quite a big ‘thing’ at present in certain circles I suspect there might even be a market for an amp with a ‘damping factor’ control!

Yes. But note that tone controls alter the relative timing behaviour as well as the amplitude. The QUAD 34 'bass lift' is a classic case of this. Using amp output impedance into the speaker impedance shifts this to knowing how to alter the amp impedance as a function of frequency. Thus if the speakers were designed using a valve amp, say, that had no feedback from its secondaries, you'd need to mimic the way the transformers impedance varies at LF if you're using a modern valve amp.
 
FWIW I've designed a few amps with variable damping factor... and it's certainly not new! RCA had a white paper about it in 1959 and several valve amp era power amps had variable damping factor and could go to infinite damping and even negative impedance... now't new under the sun.

IMO all amps should be designed for negligible output impedance, as they almost always are... The low damping factors of most valve amps are generally not intentional...
 
Yes. But note that tone controls alter the relative timing behaviour as well as the amplitude. The QUAD 34 'bass lift' is a classic case of this. Using amp output impedance into the speaker impedance shifts this to knowing how to alter the amp impedance as a function of frequency. Thus if the speakers were designed using a valve amp, say, that had no feedback from its secondaries, you'd need to mimic the way the transformers impedance varies at LF if you're using a modern valve amp.
I'm not sure I follow, are you saying that tone controls work by varying the amp's output impedance at their centre frequencies?

I am aware that using a tone control or EQ to alter amplitude frequency response causes 'phase shift' at said frequencies but I've never really understood the consequences. E.g. if I reduce the midrange output of my speakers by a -2dB with EQ, that's precisely what I hear at my listening seat, a -2dB cut in the midrange, I don't notice any audible issues with phase (or at least what my ears associate with phase).

Regarding decay times and EQ, my logic of thinking is that if you boost the amplitude response at a particular frequency, the ear will perceive the decay time of that frequency to also increase, simply because that frequency is starting from a louder SPL and therefore has a longer way to drop to below the threshold of our hearing. And vice versa, by reducing the amplitude response at a particular frequency, the ear will perceive the decay time to shorten because it starts from a lower initial SPL and therefore drops below the threshold of hearing sooner. But in a technical sense, the decay rate hasn't changed. Does that make sense?
 
I'm not sure I follow, are you saying that tone controls work by varying the amp's output impedance at their centre frequencies?

I am aware that using a tone control or EQ to alter amplitude frequency response causes 'phase shift' at said frequencies but I've never really understood the consequences. E.g. if I reduce the midrange output of my speakers by a -2dB with EQ, that's precisely what I hear at my listening seat, a -2dB cut in the midrange, I don't notice any audible issues with phase (or at least what my ears associate with phase).

Regarding decay times and EQ, my logic of thinking is that if you boost the amplitude response at a particular frequency, the ear will perceive the decay time of that frequency to also increase, simply because that frequency is starting from a louder SPL and therefore has a longer way to drop to below the threshold of our hearing. But in a technical sense, the decay rate hasn't changed. Does that make sense?

Some lower price Japanese amps in the past tended to have tone controls in the NFB loop of the power amp section :eek: and so probably would have some degree of effect on this... a very bad idea! generally tone controls don't effect output impedance though.
'fraid I ain't even going there on the damping and decay stuff you seem to have a bee in your bonnet about... it's way more down to the speaker than the amp and virtually impossible to "pin down"...
 
Valve amplifier damping factor is not constant with frequency. Typically it is much higher at the frequency extremes as the NFB runs out.
The top end only affects speakers with a diving treble impedance like ESLs, acting as a treble cut. The bass is more serious and depend on the details of the speaker tuning, but I don't think that it is a coincidence that most speakers in the valve era were sealed boxes
 
...but I don't think that it is a coincidence that most speakers in the valve era were sealed boxes

I don’t think they were, it wasn’t until AR came along until infinite baffles really took off, though the majority of speakers had ‘hard-edge’ driver surrounds rather than compliant foam or rubber along with much larger drivers that just don’t need to move as far to generate bass. Think Tannoy, Altec, Klipsch etc. Obviously some different approaches e.g. Lowther, Wharfedale etc which did have more compliant surrounds, but most vintage speakers I can bring to mind were hard edge, and it is these that need a period-appropriate amp IME. They just don’t sound right on the end of an amp from the recent ‘move a tiny low-impedance plastic driver half an inch or more’ era!
 
I suppose most domestic speakers were really variations on open baffle. I am not sure what the professional (cinema) market did back in those days, horn loading was popular
 
There is a team of German speakerbuilders who have empirically come up with a handle on the interdependence of amplifier damping factor, room volume, speaker driver and box volume. They call it Resonanztheorie.

They have found that there is a sweet spot of box volume which, when found out by trial and error, allows the driver to play best, so that the sound can really just be there, not adhering to the speaker box anymore, and without harshness etc.

Their method is to add/subtract small wooden cubes to the inside until everything locks in.

It's explained here in detail, unfortunately only in German:
http://www.hornlautsprecher.de/kapitel201.htm

Google translates almost intelligibly:
http://translate.google.com/transla...cher.de/kapitel200.htm&hl=en&safe=strict&sa=G

Having an amplifier with continously variable damping factor might make things easier.
 
I'm not sure I follow, are you saying that tone controls work by varying the amp's output impedance at their centre frequencies?

I am aware that using a tone control or EQ to alter amplitude frequency response causes 'phase shift' at said frequencies but I've never really understood the consequences. E.g. if I reduce the midrange output of my speakers by a -2dB with EQ, that's precisely what I hear at my listening seat, a -2dB cut in the midrange, I don't notice any audible issues with phase (or at least what my ears associate with phase).

Regarding decay times and EQ, my logic of thinking is that if you boost the amplitude response at a particular frequency, the ear will perceive the decay time of that frequency to also increase, simply because that frequency is starting from a louder SPL and therefore has a longer way to drop to below the threshold of our hearing. And vice versa, by reducing the amplitude response at a particular frequency, the ear will perceive the decay time to shorten because it starts from a lower initial SPL and therefore drops below the threshold of hearing sooner. But in a technical sense, the decay rate hasn't changed. Does that make sense?


I was saying that to me tone controls make more sense that trying to fiddle with the amp-speaker impedance interaction. Particulary because the music will vary as well. At LF it can make things clearer to think in terms of time delay/advance vs frequency rather than phase as the times involved get pretty large in some cases. At mid to high frequencies this all gets scrambled by room reflections.

Hard to say about the effects at LF because the room also comes into this. Hence again my preference for adjustable tone controls, not having using experiment with amp output impedance.
 


advertisement


Back
Top