I have repaired and built tube preamps and amps for years, but recently lost interest. Here is my take on tube sound:
1) Poorly designed tube preamp circuits definitely have a sound of their own. A good example is the Conrad Johnson PV12L - the circuit looks like it was designed around 12AX7 tubes and they simply changed to 12AU7 tubes to reduce the gain. As a result, the tubes run at a very low bias point on a very non-linear portion of their transconductance curve. This would color the sound for sure.
2) A lot of tube preamps oscillate when certain tubes are used (such as the one mentioned above) due to the manufacturer either not knowing better or saving a buck by not using grid stopper resistors - this would cause obvious differences in the sound signature of certain tubes.
3) I believe that MOST of the tube sound signature that people are chasing is from the interaction between the relatively high output impedance plus non-linearities of the transformer coupled output with the speakers. In fact, I don't think I've ever seen measurements where the use of the 4 ohm tap (lower impedance) does not improve the performance with 8 ohm speakers. For a lot of speakers, this interaction causes an increase in output of up to a few dB in the area where speaker impedance is highest, often in the "warmth" area of the bass region. I have only played with OTL amps for headphone use and they have similar issues, but in their case the output impedance is even higher, but without the transformer non-linearities. Oddly enough, or not, depending on your stance, feedback improves the situation in transformer-based amps - but then we have manufacturers offering "features" like a triode/ultralinear switch which does not optimize the feedback loop for one of the modes, causing weird frequency response anomolies as a result in one of the modes that seem to be pleasing to certain ears. Funny enough, it is the cost no object zero feedback designs that seem to be most popular, even though they suffer from limited bandwidth and high distortion, especially in the bass.
4) Well designed tube preamp circuits will sound very much like solid state with a slightly higher noise floor. Tube amplifiers, on the other hand, always sound like tube amplifiers due to all the compromises in trying to couple a high impedance device (a tube) to a low impedance transducer (speaker).
5) Cathode followers or buffers, which have a gain of slightly less than 1, do not exhibit "tube sound" as they do not provide the harmonic distortion profile that tube gain stages exhibit. I'd go so far as to say that any significant change they make to the sound is due to impedance matching issues or worn out tubes. Adding them to solid state devices, like DAC's, simply adds hum and noise.
There ya go
1) Poorly designed tube preamp circuits definitely have a sound of their own. A good example is the Conrad Johnson PV12L - the circuit looks like it was designed around 12AX7 tubes and they simply changed to 12AU7 tubes to reduce the gain. As a result, the tubes run at a very low bias point on a very non-linear portion of their transconductance curve. This would color the sound for sure.
2) A lot of tube preamps oscillate when certain tubes are used (such as the one mentioned above) due to the manufacturer either not knowing better or saving a buck by not using grid stopper resistors - this would cause obvious differences in the sound signature of certain tubes.
3) I believe that MOST of the tube sound signature that people are chasing is from the interaction between the relatively high output impedance plus non-linearities of the transformer coupled output with the speakers. In fact, I don't think I've ever seen measurements where the use of the 4 ohm tap (lower impedance) does not improve the performance with 8 ohm speakers. For a lot of speakers, this interaction causes an increase in output of up to a few dB in the area where speaker impedance is highest, often in the "warmth" area of the bass region. I have only played with OTL amps for headphone use and they have similar issues, but in their case the output impedance is even higher, but without the transformer non-linearities. Oddly enough, or not, depending on your stance, feedback improves the situation in transformer-based amps - but then we have manufacturers offering "features" like a triode/ultralinear switch which does not optimize the feedback loop for one of the modes, causing weird frequency response anomolies as a result in one of the modes that seem to be pleasing to certain ears. Funny enough, it is the cost no object zero feedback designs that seem to be most popular, even though they suffer from limited bandwidth and high distortion, especially in the bass.
4) Well designed tube preamp circuits will sound very much like solid state with a slightly higher noise floor. Tube amplifiers, on the other hand, always sound like tube amplifiers due to all the compromises in trying to couple a high impedance device (a tube) to a low impedance transducer (speaker).
5) Cathode followers or buffers, which have a gain of slightly less than 1, do not exhibit "tube sound" as they do not provide the harmonic distortion profile that tube gain stages exhibit. I'd go so far as to say that any significant change they make to the sound is due to impedance matching issues or worn out tubes. Adding them to solid state devices, like DAC's, simply adds hum and noise.
There ya go