On the 627 thing, I used to use these for scanning tunneling microscopy; there was not much better out there for working with signals on the edge of existence. IMO, the input current and general behavior of the input stage is extraordinarily good.
That said, I have little opinion on their sound qualities one way or another for audio use, but my feeling is that people are paying a huge premium for characteristics that probably don't matter much for audio applications. It's another case of, "It's expensive and hard to get, so it must be better."
I still prefer the newer National (TI) parts because they're inexpensive, have lower distortion and are indistinguishable if the circuit is properly designed. If you're hearing differences between high performance opamps that don't show up with spectral analysis, differential or other methods, something is wrong.
That said, I have little opinion on their sound qualities one way or another for audio use, but my feeling is that people are paying a huge premium for characteristics that probably don't matter much for audio applications. It's another case of, "It's expensive and hard to get, so it must be better."
I still prefer the newer National (TI) parts because they're inexpensive, have lower distortion and are indistinguishable if the circuit is properly designed. If you're hearing differences between high performance opamps that don't show up with spectral analysis, differential or other methods, something is wrong.
Depends on the design. You can design either to be colored or either to be neutral.
Well considering that most tube amplifiers are power rated @1% thd and most SS at .01% from a pure technical position (taste's aside) it would appear tubes have more inherent distortion , more slewing, low damping and limited bandwidth (especially in the bass).
With all due respect, i find it interesting the cognitive dissonace of those building and designing tubes, where it does appear a lot of sound vs measurements is being practiced, especially when one consider their technical stance in other areas and the importance placed on such measurement for deciding good , better , best...
Just my Observance ....
With all due respect, i find it interesting the cognitive dissonace of those building and designing tubes, where it does appear a lot of sound vs measurements is being practiced, especially when one consider their technical stance in other areas and the importance placed on such measurement for deciding good , better , best...
Just my Observance ....
Absolutely not. The right answer is: tube amps even without heroic efforts to minimize THD and maximize DF can sound better.
However, high output resistance of tubes and low output resistance of dynamic speakers require output transformers that are tricky beasts. You can get wider bandwidth, better DF, lower THD, using much more expensive output transformers, but it does not worth it, since sound quality improvement from such heroic efforts is questionable.
I myself had such belief when was young, that absence of output transformers, availability of complementary devices, guarantees better results with transistor amps. No matter how good my complementary totally symmetric SS amps were measured, anyway that stubborn professional tube beast Tesla Mono-130 sounded better, cleaner, more transparent. Much later, when I returned back to audio design, and made tube amp that measures like best transistor amps in terms of THD, bandwidth, and DF, I finally realized how wrong I was thinking that tubes were worse devices than transistors.
Last edited:
Honestly, I don't hear any magic in McIntosh solid state amps. I've repaired recently one, it has just optimal design, unlike other amps that were made in order to get some specific measurements as best as possible without sound quality in mind.
I wonder, with the basic tendency of a tube to act like a compressor more than a transistor does, maybe one raises the average level a little and the perception is more detail? Dependent on where the tube is biased along the load line of course. I don't recall much discussion about gain being non linear and its effect on subjective results. I remember back when our systems were full of peak unlimiters and various other boxes which all generated "impact" but in the long run were dismissed as not as musical. An LP at 65 dB range on a good day sure does not have the theoretical level of detail a CD does, but if you don't mind the other aspects, can be more, dare I use such a subjective term, involving? This should be easy to test. I think one of my DSP packages has a compressor.
Grasping at another straw, we typically measure amp distortion with a resistive load. Should we be measuring the difference to a captured sound from a reference speaker system? Are some of the subjective differences only measurable with the complex loads, or as Linkawitz describes, totally evil passive crossovers? Some far more evil than others. I might expect the full range driver camp to chime in here as that is the basis for their pursuits.
I quit playing with tubes after a few rounds of adding solid state rectification, bias control, and phase splitter. Each both measurable and sonicaly better. The end game would have to have been replacing the outputs with transistors. End of my tube experiments.
Grasping at another straw, we typically measure amp distortion with a resistive load. Should we be measuring the difference to a captured sound from a reference speaker system? Are some of the subjective differences only measurable with the complex loads, or as Linkawitz describes, totally evil passive crossovers? Some far more evil than others. I might expect the full range driver camp to chime in here as that is the basis for their pursuits.
I quit playing with tubes after a few rounds of adding solid state rectification, bias control, and phase splitter. Each both measurable and sonicaly better. The end game would have to have been replacing the outputs with transistors. End of my tube experiments.
This , is unrealisticaly exagerated and at odd with all known
measurements/comparisons of best tube amps vs common SS amps,
let alone these latters best ones , unless you re talking about the
early transistorized amps , of course...
Last edited:
I'm mystified as to what the price range has to do with it. I would expect the difference between good and bad to be in terms of design, not the cost of the hardware for a given output power, for example. I have read some threads where a design is improved (measurably) by removing an end of a capacitor from one point, and re-connecting it somewhere else, or changing the values of a couple of resistors. Even introducing extra components must add only a few cents or a few dollars at most to the price, surely.
This is what I would have thought myself, yet would tubes make for a better VAS, stage with mosfets on the current side .. An tube amp with no transformer ...
Anyway the measurements of tube amplifiers , especially on simulated speaker load throws a monkey wrench in whats sounds good vs specs.....
TVR-Geek, Please don't buy an HCA-800 and hope to come back to me with what you hear or measure. It is listed below my acceptance level, and may not even have any circuit topology that I am generally familiar with. I would recommend an HCA-1000, 1200, 1500, or 2200, if you want to buy a used Parasound Amp. At least you will have something then that we can talk about, as I am associated with those designs. The others may be OK, but they are completely outside my control, so don't complain if you are disappointed, and they are generally designed for the absolute minimum in heatsinks, parts and materials that they can get away with.
The same can be done with transistors, but you will need much more sophisticated topology than couple of tubes. And of course, due to many devices involved result will be worse.
What I do in tube power amps, I sense screen grid current of output tubes and control by this result an opto - attenuator. It works exactly like an optical compressor that engages as soon as the amp approaches clipping.
You may ask SY, we measured my Pyramid on his laboratory setup.
I have no intentions to exaggerate, when I work to satisfy my own curiosity. And I always find that common beliefs, fashions, and legends are plainly wrong. I was one of victims of them before, when design of electronics equipment was my profession.
This is what I would have thought myself, yet would tubes make for a better VAS, stage with mosfets on the current side .. An tube amp with no transformer ...
Such divisions on VAS, current stages, etc... is not the best approach to build the best audio device. However, amps with vacuum tube preamp and mosfet outputs sound very nice. The same with measurements.
Tried one of the big stereo halo last year, good sounding amp but the protection would interrupt us after about 10 min of playing....
interesting to hear in what way do tubes measure superior to SS and what in those measurements make you know its a good sounding amp..?
Dynamics. Dependence of distortions on frequency, on power output, phase - related distortions. Tubes don't have Kirk and Early effects, don't change base width, don't have such thermal distortions. They don't care what frequency in audio band how to distort. And what is most significant, it is easier to design tube amps such a way that distortions depend on power such a way they are less audible. SY saw how my Pyramid behaved, on 40W of output power 2'nd harmonic was visible at -80 dB, the rest was below noise floor. The less was power, the less was that 2'nd harmonic. However, approaching full 80W the amp started showing growing up grass of harmonics, but such behavior is not audible as distortions, due to psycho-acoustical properties of perceptions. Also, dumping factor was high (though, I did not bother to measure it) due to nested feedback, part of which was directly across output tubes.
Such experiments show that despite of very low level of measurable distortions tube amps still sound as tube amps: clean and transparent. That means that myth about "people like tube distortions" have absolutely no foundation under it.
However, I took care of clipping adding compressor-limiter, because clipping in any amp with deep negative feedback, be it on diamonds or steam engines, sounds as clipping, and is unacceptable to my taste.
Last edited:
- Status
- Not open for further replies.