Whaaaat? Dont like that one?
Different, isnt it? 🙂
Well, that was my take away from that time period (more like 25 years ago).
Different, isnt it? 🙂
Well, that was my take away from that time period (more like 25 years ago).
Last edited:
You can tell this story to those who did not do homework. 😀
35 years ago I played a lot in a laboratory designing analog synthesizers, guitar effects, and hi-fi amps. My experience tells opposite. 😀
Some people might like some forms of distortion, but I doubt if anyone really likes crossover distortion!
You were there, I mentioned SET amps and they were all over me. BTW I might point out that most of these capacitor, resistor, etc. distortions tend to both be of low order and decrease to zero as the level goes to zero, just as those beloved tube amps, LP's, speakers, etc. I still need to be shown a cap that even approaches the temporal and harmonic artifacts of even the best speakers.
Well Richard, I have not heard this effect, but I believe that it is possible. But then, I have had similar experiences that have been 'put-down' by the same critics.
Some years ago, I heard a Wavac SET demo that 'knocked my socks off' at CES (The Show) and yet the measured distortion as shown by John Atkinson was measured as being 'off the map' for that design. Go figure! But I stand by what I heard.
Some years ago, I heard a Wavac SET demo that 'knocked my socks off' at CES (The Show) and yet the measured distortion as shown by John Atkinson was measured as being 'off the map' for that design. Go figure! But I stand by what I heard.
Another one of those effects that decades from now many will wish that it can be un-done from the final mix.
Class B is more dynamic than AB because the PS is less
constantly loaded.
In some situations that might swamp out the crossover.
I've found one secret: in order to make class AB amp sounding like SE class A amp it needs to have power supply of the same power as would be needed for class A amp. I.e. designed for constant peak power delivery. Such a way dynamic is very nice, and class AB amp can compete with SE class A amp, and be even better for orchestra reproduction.
Sorry, I have no idea what you are saying. Could you expand your comment? You appear to be saying that Class B loads a poor PSU less than Class AB, so that crossover distortion is swamped out by what?hitsware said:
There might not be any such thing...? The test I did makes me think affects of DA can be audible and thats about it. And --
I still need to be shown why a phono cartridge - with its limited, non-flat freq response, very high thd and IM, limited channel balance and crosstalk, and all sorts of artifacts -- let alone turn table artifacts, LP wear and ever higher distortion and noise with each playing, - pre and pre-pre's, tone arms et al- seems to be accepted as an accurate source (wtf ?)
Its was never accurate IMO. Like and enjoyable, perhaps. But, I think that would better be ascribed to the distortion (linear and nonlinear) being liked and its affect. -RNM
Last edited:
Accurate in terms of hearing and measuring are different things. It is obvious. You measure distortions, minimize what does not sound like distortions, minimizing that add what sounds like distortions, but not measured. You get better measured results, but hear more distortions.
What you call "Enjoyable" I call "Ignorable". This is the main difference.
THD/IMD measurements fit to visual perception (charts, graphs, diagrams), but not to hearing perception. We have a small problem then 🙂
Not 'swamped out' on a scope, but perceptually 'swamped out'
I.E. If you had a bias knob you could easily adjust, you may find
that certain music, speaker location, etc. would sound better
with a lower bias setting.
There are two bias settings which work: the 'correct' setting (which may be surprisingly low for some output topologies, such as CFP), and the 'first watt' setting' (much higher bias so the switchover kink is moved well away from the zero-crossing region). Anything less than the 'correct' setting is unlikely to sound nice because the feedback loop loses gain just at the point it needs it most.
A 'lower' bias setting may be OK for those people who normally set it too high, as being between 'correct' and 'first watt' bias may be worse than either of them.
A 'lower' bias setting may be OK for those people who normally set it too high, as being between 'correct' and 'first watt' bias may be worse than either of them.
When THD on nominal power is -80 dB of second order only, and goes down on less power, it is ignorable even if it goes up to -20 dB on peak power. But when it stays around -80 dB on any level that means something is wrong. Things are even worse when on power close to peak power THD is -100 dB, but goes up on lower power. Even when people say, "Nah, it is THD+N, it can't be heard on noise background", it is wrong. Distortions MUST go down with lower power, their order MUST go down with lower power, otherwise they can't be ignored by ears that were trained during long period of evolution to ignore such distortions.
One more source of audible distortions is their burst on fast decay (or attack) of sound level, while they are low on steady measurements. Here capacitors play the major role, also charges in semiconductors and thermal effects.
One more source of audible distortions is phase modulations of signals by itself. Yes, we can't hear sinewave sounds with period of less than 50 microseconds, but phase shifts 10 microseconds and less are audible when highs are modulated by phase by lows.
Different topologies have different illnesses, but what is remarkable, a single ended triode tube stage has less of them. Trying to "improve" this stage you can see how this 3 illnesses are getting worse, if you ignore them in order to get something else better.
One more source of audible distortions is their burst on fast decay (or attack) of sound level, while they are low on steady measurements. Here capacitors play the major role, also charges in semiconductors and thermal effects.
One more source of audible distortions is phase modulations of signals by itself. Yes, we can't hear sinewave sounds with period of less than 50 microseconds, but phase shifts 10 microseconds and less are audible when highs are modulated by phase by lows.
Different topologies have different illnesses, but what is remarkable, a single ended triode tube stage has less of them. Trying to "improve" this stage you can see how this 3 illnesses are getting worse, if you ignore them in order to get something else better.
So we go full circle, the reasons for the audible differences are not measurable.😱
Your full circle within 24 hours with different wording. What more can one wish?
Scott. Despite of this:
Can you please tell me if the doped silicon -N, P lands- in semiconductors (for audio at least) can have piezoelectric properties?
Thanks
Regardless of my assumptions, your post made me search and find that high k ceramic capacitors -Class II & III -exhibit some short of behavior (in addition to piezoelectricity) that I was not aware of:
Dielectric constant/ac voltage nonlinearity and hysteresis due to ferroelectricity.
Strong reduction of capacitance with increase of electric field strength (dc bias +ac signal amplitude).
Reduction of capacitance with temperature increase.
Rapid aging (reduction of capacitance with time).
I have to thank you for that. 🙂
I still think that Heaviside was theorizing about two formed/fixed capacitors having dielectrics with different e, all else equal. George
- Status
- Not open for further replies.