With all the zero feedback valve amps there is a way easier explanation - though no one seems to love it.
In old days when movies did have an optical soundtrack it was well known to the experienced ones, that no small part of noise and ringing into sound was introduce not only by (that times) pretty limited S/N figures of the electronics but also by mere pickup of mechanical noise of the projector.
This was the case due to the mic sensitivity of the tungsten bulbs illuminating the optical sound track.
So - the simple explanation for those direct heated triode sound is - apart from its transfer function - a buried function like that of a spring reverberation gear.
All those fancy silicone rings applied to tubes do speak for themselves - also - any serious listener should have made the experience that some tube amps can sound "glassy" right away - nothing else than a sonic pattern overlay by the materials in use.
Michael
In old days when movies did have an optical soundtrack it was well known to the experienced ones, that no small part of noise and ringing into sound was introduce not only by (that times) pretty limited S/N figures of the electronics but also by mere pickup of mechanical noise of the projector.
This was the case due to the mic sensitivity of the tungsten bulbs illuminating the optical sound track.
So - the simple explanation for those direct heated triode sound is - apart from its transfer function - a buried function like that of a spring reverberation gear.
All those fancy silicone rings applied to tubes do speak for themselves - also - any serious listener should have made the experience that some tube amps can sound "glassy" right away - nothing else than a sonic pattern overlay by the materials in use.
Michael
Hi,
definitions should be explicit, concise and universal. I`m not able to come up with a satisfactory formulation.
Slew rate designates a dynamic range given by the maximum rate of change of a signal at some point in an amplifier (input or output). Also, I see many possible causes of limitation.
definitions should be explicit, concise and universal. I`m not able to come up with a satisfactory formulation.
Slew rate designates a dynamic range given by the maximum rate of change of a signal at some point in an amplifier (input or output). Also, I see many possible causes of limitation.
Wu,
The output voltage is indeed a signal, but you are missing the point. The slew rate of an amplifier is the MAXIMUM rate of change at the output. It is not a property of any signal in the amplifier. Voltage rate of change at any given point on a waveform IS a property of a signal.
Cheers,
Bob
I agree. Maybe I should have written "clipping".
The reason of my initial question : there are a surprising number of CD records which contain very short durations at Full Scale Deviation which correspond to clipping. They are almost undetectable.
I wonder if distorsion having slew-rate limits as origin can be considered as clipping and if we really often hear it.
Pavel,
Here I think you are saying, and I think you are touching on two different things in your two sentences.
In your first sentence, you are saying that if you increase negative feedback to reduce 2nd and 3rd, but 5th and/or 7th go up, that is a bad thing. I agree. I also point out that such a situation is NOT generally the result in spite of what some people say about the Baxandall results.
In your second sentence, you appear to be saying that, even absent any increase in the 5th and/or 7th, it is bad to apply feedback that ends up decreasing the 2nd and 3rd.
In other words, you are arguing about the spectral distribution and balance of the harmonics, even apart from the absolute values. Put another way, if you have two amplifiers both with 5th and 7th at -110 dB, but the first has 2nd and 3rd at -90dB while the second has 2nd and 3rd at -110 dB, then you are saying that the second amplifier sounds worse (because it has less 2nd and 3rd). This is the old "distortion masking" argument which is highly controversial. The notion that more distortion at lower orders somehow masks the nasties of very small amounts of distortions at higher orders is questionable, but not impossible.
In this context, you must realize that what we hear as the nasties of distortion is NOT usually harmonics, but rather intermodulation products. Many people miss this point. The spit you sometimes hear on a cymbal that is played through a system with distortion is not ultrasonic harmonics but rather intermod products of the cymbal frequencies that are reflected down into the more audible portion of the audio range. It is easy to begin to see this if you look at the spectrum of a 19+20kHz CCIF test tone fed through an amplifier. The 3rd lies at 18kHz, the 5th lies at 17kHz, the 7th lies at 16kHz and so on.
There is another very important point here in distinguishing between harmonic distortion spectra and IM distortion spectra. We all know that negative feedback becomes less effective at higher frequencies. This means that if you started with an open-loop amplifier with equal-level harmonics from 20 kHz fundamental on up, when you put typical feedback around it, the lower harmonics will be reduced more by negative feedback than the high harmonics, possibly leading to what you are complaining about.
BUT this is not so for IM distortion in the audio band. Once again, look at a CCIF IM spectral plot. In that case, the higher amount of feedback at lower frequencies is at least as effective for high-order IM products as for low-order products (at least for those reflected further down into the audio band).
Two very different animals, and it is important to distinguish them.
Cheers,
Bob
If I may split some hairs: if you want to test for max slew rate, just increase the slew rate of the signal you send into the amp until you see that the amp becomes non-linear. At that point, you know the amps slew rate limit (which at that point is the slew rate of the signal you send into the amp).
If you would send in a signal with a very fast step you surely will put the amp into a slew-limited mode, but you still don't know what its max slew rate actually is. You only know that you exceeded it.
I also must agree to Tom that 'slew rate' is a property of something that moves with a certain rate of change. A step signal with a peak-to-peak level of 10V and a risetime of 1uS has, obviously, a slew rate of 10V/uS. A 10V peak-to-peak 20kHz sine wave has a max slew rate of 1.25V/uS. That max slew rate happens at the zero crossing of that sine wave.
Bob referred to anti aircraft guns; a gun turret that slews (yes that it the correct word) a full circle in 4 seconds has a slew rate of 90 degrees / Second. Except that in the army, a circle is divided not in degrees but mills, where a circle consists of 6400 mills. So that gun had a slew rate of 1600 mills/ Second. And so it goes.
jan didden
Electron mobility issues come into play at much higher frequencies, unless by 'electron mobility' you mean something quite different from the usual use of the term. In that case we can't have a conversation without someone to translate for us.
Stability does not limit dynamic range. Stability limits allowable feedback.
Amplifier 'speed' is an almost meaningless term. It mainly seems to be used by people who don't understand amplifiers.
I too am called Elvis!
Stability does not limit dynamic range. Stability limits allowable feedback.
Amplifier 'speed' is an almost meaningless term. It mainly seems to be used by people who don't understand amplifiers.
I too am called Elvis!
Hi Bob,
I do not think so. Maybe it is a result of my poor English. What I mean is that even if I measure very low overall THD (like 0.003%, e.g.), I do not recommend a situation when 5th and higher harmonics have higher amplitude than 2nd or 3rd harmonic. In other words, I prefer decaying amplitude of spectral components (with frequency) even if the resulting THD is very low. I do not like to see 2nd harmonic of -118dB and 5th harmonic of -105dB, as an example. That is what I intended to say, probably I was not very clear.
Regards,
DF96,
By definition, electron mobility signifies the speed of electron movement (at all frequencies).
velocity = displacement per time interval, the magnitude of velocity is speed, that is change of position per time interval.
Instability due to high open-loop gain and massive phase errors is the main reason for the typically pitiable frequency range (i.e.dynamic range) of global feedback amplifiers.
By definition, electron mobility signifies the speed of electron movement (at all frequencies).
velocity = displacement per time interval, the magnitude of velocity is speed, that is change of position per time interval.
Instability due to high open-loop gain and massive phase errors is the main reason for the typically pitiable frequency range (i.e.dynamic range) of global feedback amplifiers.
No, electron mobility is a term for how fast an electron can move through a particular crystal lattice. As I said, if we are speaking different languages then we can't have a conversation.
You now seem to be confusing dynamic range and frequency bandwidth. Do words have different meanings on your planet?
In Alice In Wonderland words mean whatever the speaker wishes them to mean. Not true here on DIYaudio.
You now seem to be confusing dynamic range and frequency bandwidth. Do words have different meanings on your planet?
In Alice In Wonderland words mean whatever the speaker wishes them to mean. Not true here on DIYaudio.
Last edited:
You seem to have missed the fact that global feedback increases the frequency range of an amplifier. Which, of course, has nothing to do with dynamic range.
Elvis has left the building
jan didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.