Tube amps are not technically as accurate as solid-state but they sound more musical, I would submit that they sound that way because of the ring of the tubes just like the reverb of the Stradivarius violin. I believe the vibration of the sound from the speakers excite the tubes and there is a pleasant reverb effect. In mixing vocals there is an important effect in the reverb processor called pre delay and that time delay before the reverb is actuated in the processor is like the time delay of the speakers making the tubes ring. Thoughts?
You are right, a designer must use feedback that is always a compromise but transistors are still more accurate (don’t sound better) than tubes if you look at them on a scope. Another fare worse aspect is how transistors distort when clipped they sound awful but tubes sound smooth and forgiving. Audiophiles use underpowered amps in general, it takes a lot of power to produce loud sounds from some of our big not so efficient speakers. Tube amps are always way lower power than ss amps but because of the nice distortion it doesn’t matter as much. But physics is physics it takes a certon amount of power to produce a high volume from the speaker and it is very rare that tube amps have even close to enough power in fact it is the exception that even ss amps have enough power to produce movie theater volumes at home. By the way even golden ears have a hard time hearing <15% distortion at high volumes.
@donavabdear Most of the last quote is incorrect.
Its more than just looking at the signal on an oscilloscope. The ’scope won’t show you the harmonic content very well at all, and of course its harmonics that cause all instruments to sound the way they do.
Its harmonic distortion and IMD that causes amplifiers to sound the way they do. Based on your comments, my surmise is that the solid state amps you’ve heard don’t have enough feedback. Plus what feedback they do have (which might be 25dB, not nearly enough) was applied at a non-linear spot near the input of the amp (usually a transistor) which caused the feedback signal to be distorted before it could do its job. As a result of the non-linearity, higher ordered harmonics were added to the resulting sauce.
Human ears complicate this in two ways. The first is that the ear uses higher ordered harmonics to sense sound pressure so is keenly sensitive to them (the ear has about 130dB range)! The second problem is the ear converts all forms of distortion to tonality, just as it converts the harmonics of musical instruments to tonality. So you can hear the distortion most solid state amps make quite easily- as harshness and brightness. Its a coloration every bit as much as the ’warmth’ of tubes caused by the 2nd and 3rd harmonics!
The nice thing about the 2nd and 3rd harmonics is they can mask the presence of higher ordered harmonics. This is the sole reason why tubes sound smoother than solid state.
Tubes have traditionally been better at low level detail such as decay of musical notes in a room. This is because the phase of the 2nd and 3rd harmonics help the ear to hear details like this, which includes the 3D sound stage.
It is possible to build solid state amps that sound like tube amps. All you have to do is have the same relationships of lower ordered harmonics to the higher orders (and not get messed up with IMD in the process). This is really hard to do with traditional class A or AB solid state amps, but IME it is possible to do with class D. What I am saying is there are class D amps now that sound like a tube amp- the best way to tell the difference is the class D is slightly better in the detail department.
WRT feedback: Most amps don’t use enough! This is because of two reasons, the first being that adding more feedback might cause the amp to become unstable- that it is more likely to go into oscillation when the chips are down. The second is the amp needs a lot of something called Gain Bandwidth Product, which as you may have surmised is a mathematical number based on how much gain and bandwidth the amp has. If the GBP is too low, feedback will decrease with frequency, meaning distortion will increase with frequency. This is a recipe for brightness and harshness- a ’sheen’ on top of the strings, stuff like that.
All tube amps lack both the phase margins (that oscillation issue) and the GBP to use enough feedback. But they have an advantage: their feedback node is usually a bit more linear than that used in most solid state amps, so the feedback signal gets less distorted, and so does less damage while otherwise doing its job. IOW: smoother sound because less higher ordered harmonics are generated.
Most solid state amps have the same issues I mentioned in the last paragraph, but they usually do a bit better so more feedback can be applied. But as I mentioned the feedback signal is also more distorted so the feedback does more damage doing its job.
In a class D amp you can run a whole lot more feedback and can get around these problems, depending on the topology. As a result there are some class D amps that are as smooth as the best tube amps but more transparent since they have less distortion overall. But like tube amps, not all class D amps are equal!!
