Fully agree! This is just one example demonstrating the overall superiority of the high impedance drive; there are also many other aspects of it. High damping factor is good for only one thing, damping the main speaker resonance. Elsewhere in the audio band, low impedance drive increases distortion.
it is a matter of finding the right impedance, not deliberately increasing impedance.
There is so many audio authors who concluded by math formulas that damping factor actually doesn't affect much the speaker resonance because it is diluted.
A damping factor of 2 would result in greater frequency response variations and is a miss for the deep bass and complex crossover control.
I don't think df is the reason for tube amplifiers to sound better.
WRT Damping Factor, the specification as cited by manufacturers is as tested at 1 Khz; it falls as you move lower into the frequency band. You cannot rely on a manufacturer's Damping Factor specification since we want the high DF at low frequencies, say, at 160 Hz and below to perhaps 30 Hz or the standard 20 Hz if you prefer.
DF as a factor in a sound system includes both runs of speaker cable plus the crossover and the drivers themselves. All these factors reduce the effective DF versus using just the loudspeaker impedance alone, so once more we have a specification that cannot be relied upon without specific system measurements at the frequencies we are interested in.
In order to increase the DF, the amplifier designer only has to increase Global Negative Feedback, an exercise that improves the specifications while degrading the Sound Quality.
So, I am inclined to dismiss Damping Factor as a worthwhile criteria in amplifier selection.
DF as a factor in a sound system includes both runs of speaker cable plus the crossover and the drivers themselves. All these factors reduce the effective DF versus using just the loudspeaker impedance alone, so once more we have a specification that cannot be relied upon without specific system measurements at the frequencies we are interested in.
In order to increase the DF, the amplifier designer only has to increase Global Negative Feedback, an exercise that improves the specifications while degrading the Sound Quality.
So, I am inclined to dismiss Damping Factor as a worthwhile criteria in amplifier selection.
1. Positive feedback by current also increases damping factor,
2. Negative feedback can be nested, not necessary Global only,
3. It is a myth with thin foundation (valid in some cases only) that GNFb degrades sound quality.
I am pleased to hear someone saying this!
After becoming re-interested in audio only recently, I was shocked to find several brands and even self-builders happily proclaiming the use of DSPs in a speaker setup as if it was the new and fashionable thing to do. My first reaction was that that can't be a good thing, but I put it down to me being behind the times...
Nice to hear it's not just me after all.
We are not so far from each other. But I disagree that lots of information hasn't made it to the books. And your example of a capacitor doesn't work. In my version of Spice I can enter a ESR value for every capacitor and if I want to I can add all parasitics known to me or that I can measure if I have that cap on my bench. I can add resistance for every wire and connect traces with capacitors that show capacitive coupling. It is as exact as you want or need it to be. In engineering that concept of parasitics and the equivalent circuit diagram are well known and available.
That's an excellent point! I feel to spec an output transformer an impedance plot of the speaker connected to it needs to be available. Jeffrey Jackson did an interesting, short blog post about damping here: hifi heroin: Optimal Source Impedance....
While there is a clear definition of damping factor and feedback and both can by quantified, "degrades sound quality" can be whatever someone wants it to be.
Damping factor and other means are needed to get sound quality that means the end result, when sounds seems as real as possible. When subconscious reactions on sounds happen before you realise that sounds are reproduced by electronics.
No.JonSnell Electronic said:
Most valve amplifiers, like most amplifiers, produce second (an even harmonic) as the dominant one.
In any new area of learning, it can be difficult to tell the difference between a debate between different schools of thought and an attempt at teaching a reluctant class who are confused about the basics. Many of the 'debates' on this forum are actually attempts at teaching - which usually fail.shakeshuck said:
You need to gain understanding and then make up your own mind. You also need to be clear what you want: faithful sound reproduction or a sound which pleases your ears.
Start reading textbooks. Start building circuits. Modify them. Redesign. Listen. Measure. Repeat. Don't believe everything others tell you.
Valve circuits can be designed for minimum distortion, and are then virtually indistinguishable from solid-state. They can be designed for more distortion, which is moving away from hi-fi, and some prefer this. There is no such thing as 'tube sound', just as there is no such thing as 'BJT sound'.
We know how to design and build amplifiers which can reproduce a music signal. This has been known for about 60 years. For some people this is enough. Others want something different (often something less, but sincerely believing that it is something more).
Do you seek understanding or simple recipes? If you want recipes, are you willing to accept other peoples' judgements on what the result tastes/sounds like?
In programming you can always call a library function, but to get the best from a library function it helps to know how it works so you don't give it invalid parameters.
DF is only relevant for the bass response, not impulse response. This is because only around the bass region is the speaker mechanically well coupled to the electrical circuit.45 said:
Most capacitors in most circuit roles cannot colour the sound, unless an entirely inappropriate dielectric or value has been chosen. Those who think of themselves as "sophisticated" find this too simple to be true so they prefer to believe something else.sser2 said:
The frequency response of a system is simply the Fourier Transform of its impulse response. So impulse response is enough to characterize a system over the full frequency range, including the bass.
Perhaps it was a bad analogy under the circumstances. Of course programming isn't subjective, unless you're arguing about tools. It works or it doesn't.
I was more bleating about how some tasks which should be relatively easy are often over-complicated, and finding the easy route is often well hidden.
Well, you cannot get away from what DF96 said.
Doing so, you'll gain precious experience that cannot be replaced by random forum posts. And of course, please stay safe.
First that driver is not my choice because I don't like fullrange drivers with whizzer cone. The point is just that generalizations simply don't work for best results. Period.
The bass loading of that driver was a resonant type with actually two added resononces, one above and one below the driver resonance and yet they do not appear anywhere.
The improvent on the overshoot is just theroetical because it only shows up in anechoic conditions in a very narrow frequency range around the resonance. The tiny peak at resonance completely disappears in normal listening room. Room modes rule.
With other kinds of bass loading like sealed box or 1/4 wave transmission line one might not even get that, if properly done!
Instead the overall performance is measurably and audibly better aross 99% of its frequency range. It's not simply opinion but facts.
The in room FR clearly shows (high Zout on the right and low Zout on the left) that there is abosolutely no problem whatsover at low frequency. On the other side the low Zout doesn't solve the issues related to the whizzer cone as this is not "connected" and calls for a tweeter! Imagine the total disaster of a full range with a whizzer cone + a tweeter crossed-over at 2-5 KHz....
The 3-5 KHz peak is due to the whizzer cone. Nothing can be done about that whitout degrading overall performance.
Attachments
Last edited:
I didn't say that the driver is your choice. You said that you preferred the sound with high Z/low DF over the alternative - that was your choice.45 said:
The 3-5kHz peak seems to be more noticeable in the high Z plot because the lower frequencies are attenuated. My reading of the plots is that low Z has a better frequency balance, although both are lumpy because of the room.
If there is a cabinet resonance interacting with the speaker resonance then the coupling to the electrical end may be weaker. However, it remains true that most speakers are designed for voltage drive and most amplifiers (tube and others) are designed to provide voltage drive.
You might be the only one on Earth. The FR with the higher Zout amp is unquestionably better. It is approx. within +/- 2 db from about 45Hz to 20KHz respect to the average level, except for that peak. With low Zout there are no trebles, less lower bass, more rollor-coaster!
The entity of the 3-5KHz peak is the same, about 4dB.
For what reason low frequency should be attenuated? There is actually more output below the driver resonance which is around 60Hz (in that box)! It's the trebles that increase with Hi Zout to make it flatter. That's the effect of coil inductance....
Lumpy? That measurement is from the listening position of a typical living room.
Most tube amps do not provide voltage drive at all. Voltage drive means that they keep the voltage constant with varying load for a given input. That simply doesn't happen like for SS amp.
Most of them will behave like fairly constant power sources regardless of their Zout and so flat impedance is still desireable.....
Last edited:
Worth repeating the corrections already posted by others, push-pull amps produce odd harmonics regardless of tube or transistor.
Even harmonics are typically interpreted as pleasing to ears, not odd harmonics.
I am simply reporting what I see in your plots.45 said:
Yes, I am unique.
I assume of course good design, which means NFB which means low output impedance. Even SET with no NFB can achieve a vague approximation to voltage drive if the output valves are big enough.
No. It is Zout which determines whether an amp provides roughly constant power or not, which is the opposite of what you have just said. Low Z means constant voltage; high Z means constant current; middling Z means constant power.
- Status
- Not open for further replies.