Discussion in 'Power Amps' started by shaizada, Nov 23, 2016.
Three decades back there was Bob Carver and his challenges, which was pretty interesting stuff.
Is that an inherent property of tube amps, or just SE amps? PP tube amps have even order distortion cancellation and don't sound too tubey.
My point exactly. Distortion level + "ear friendly" harmonic distribution --> "tube sound".
As an aside to this discussion, what type of distortion is produced by guitar amps/pedals/etc? They have all sorts of tone/overdrive/saturation/etc capabilities.
I'm guessing some act as soft clip/compression, others shove in tons of harmonics... but I'm just taking shots in the dark here.
Would you agree that ''tube-y'' and ''tube sound'' can be separated?
The notion that 2-nd and 3-d harmonics generate sense of depth and expansiveness in my experience is utter BS.
Or soft-core clipping that sort of manifest in the former + some possible phase crap.
In my experience amplifiers that sin these ways only veil and/or warmfag sound, and not enhance it in any positive way. (3-rd feels like veil, 2-nd like warmfag)
I can understand why some mistaken veiled to feel like deep stage. It feels claustrophobic, like cave. Maybe also warm, as probably there's a cozy fire in the cave.
From the pov of high orders: Probably one of the cleanest SET amplifiers around, the EC Studio doesn't have high order distortion in ppm level like so many high feedback amps, and still manages to have higher resolution, better soundstage, etc. Ie tons more of relevant information. (Hundreds who have heard it are hard to ignore, unless they are all morons and sheep, me included.)
Why great tube amps do better than any SS in soundstage and many other areas? Most of us here agree on that observation, yet
nobody has clearly in technical terms been able to convince me. No measurement yet conceived correlates, let alone proves this.
One mechanism of it hard to ignore is ballistic propagation vs plasmons/e-gas. Almost any other aspect of tube and tube circuit can be emulated with silicon.
I will eat my hat made of thick cow hide once I have heard SS amp that sounds just like my EC. (the hat will survive many years, I'm committed) I mean purity of tone/timbre, unhindered dynamic scale, proper precise positioning in 3-D canvas.
If someone commercializes graphene transisor in my lifetime, I might have to eat my hat.
SE ones make even harmonics, P-P ones odd, just like regular amp, but in significant levels. Clipping and harmonics go hand-in-hand. Compression is something else.
SS guitar amps for some reason are scarce. Nasty-nasty sounding, so they say.
Usually gain is very high, low or negligible feedback for low order harmonics and 'headroom'. Pentodes will have more higher orders than triodes, but still low enough.
A friend of mine who builds these says BJT are worst of all and JFET somewhat usable.
Tom is more keen on this, I hope he can go in more detail. I'm curious too.
Would you start by defining the terms for us then?
I wouldn't suggest that anybody is an idiot. I would suggest that everybody reading this (and the subset who have heard the amp you refer to) are humans, though. As such, they'd be affected by common cognitive psychology effects, such as confirmation bias and such. This doesn't mean they're idiots. It only means they're humans, and I'd invite folks here to think of that as a possible confounding variable.
No measurement will ever prove anything. That's an impossible standard to meet. We can't even prove the existence of gravity.
Would you expand on this? I'm curious about your comparison of tubes vs semiconductors from a quantum mechanical perspective.
Actually, SE generate dominantly (but not only) even order distortion. Typically, you see at least H2, H3, and H4 from a SET. PP adds more harmonics (both even and odd, but odd tend to dominate).
Based on what? I can build a quality amp that uses BJT, MOS, JFET, and tubes, including combinations of these devices. Making a blanket statement that A > B > C is rather meaningless, in my opinion. Each device type has its advantages and disadvantages. A competent designer will be able to make the advantages dominate in the system performance.
Newton would disagree
We may indeed not yet understand fully all of its effects, the intricate mechanisms, especially in different conditions (distance, velocity, etc.), but we surely can prove its existence. If in doubt, next time outside* bring about an HP1, throw it right above your ahead as high as you can then do not move and observe where it lands... If the resulting expenditure isn't proof enough for you, I don't know what is.
Though if you're still worried about confirmation bias and stuff**, then I suggest you DBT this, varying the weight of the payloads, the height, all in completely controlled double-blind conditions so that you don't mistake gravity for placebo effects!
* Idiot's disclaimer: Do NOT do this!
** Because we're all human after all and always subject to such common cognitive psychology effects, are we not?
Notable amounts of 2-nd and 3-rd harmonics are ''tube-y''.
Well layed out triode with good OT-s is 'tube sound' to me, with reasonably low distortion vs ''tube-y'' amp.
I threw that in to be purposely preposterous, yet I feel such things have not been discussed much.
Will ballistic flow vs fluctuation of electron gas be significant in the chain where e-gas is not avoidable in form of wires? Is of course sketchy at best. The ballistic nature will have effect on switching characteristics, again probably in a diminutive scale, perhaps insignificant in audio use.
Another thing: Semiconductor lattice will interact with the signal in very small signal scale through interaction of holes and fluctuations of Fermi level, whether or not this is significant in actual application current densities, again I don't know, probably not. Different semiconductor materials and doping mechanisms have different noise characteristics on top of the usual suspects of construction, type of device and macroscopic parasitic effects. Tube with wildly different medium and semiconducting mechanism must have different nature in this also, again on top of the culminating characteristics of awesome linearity and negligible grid capacitance, parasitics etc.
All of this might be relevant after taking into account of already known stuff about circuits and their components.
I don't doubt you could build quality g-amp with any of these devices, but I do doubt that they all sound equally good.
Why are there so little SS g-amps around? Let's leave out cultism, mental bias to tube glow and etc ridiculous claims for constructive discussion.
I agree that these psyche things affect everyone, especially hi-fi nuts like folks here. I have strong bias pro tubes, I don't deny it. Why do I consider this subset of impressions as competent way in valuing gear is that I have given 'neutral' persons (my mother, friend who plays in orchestra, few other musicians who play various instruments) a listen, they seem to favor same reproduction chain as myself. This sort of validates my senses for myself, of course this is personal.
Newton observed gravity and devised math to describe observations, but he did not prove it's existence in terms of physics model through physics constants. Observation is observation, not proof.
With the advances in particle physics and increased understanding in dark matter we might see the day gravitational quanta will be detected with adequate reliability. Gravitational waves detection hint that we have the capability.
Google found me IEEE article about tubes vs solid.
Spoiler: Move along...
Not to derail, but proof comes from observation, never from theory or modelling (which can be idle or educated guesses). Just because things like the Standard Model doesn't incorporate gravity is more indicative of a fundamental weakness in the modelling framework than lack of proof for the existence of gravity. And the way I see things, 'dark matter' is an assumption to explain a contradiction between observation and current understanding. Other assumptions (e.g. TeVeS extension of gravity) exist, which don't require imagining a mysterious new kind of matter that no one can detect.
I'm not sure I'd call that a "confouding" variable. A cognitive psychological effect is as real to the subject as something more objective. If a placebo has the desired effect, and has fewer side-effects than a "real" intervention, then the placebo is better.
You know how they lengthen the sound of gunshots for movies/games? Short transients are more difficult for our brain to register. I have a hunch that tubes provide a similar thickening or stretching which makes detail more readily digestible. It makes sense to me that tubes could be superior for detail-oriented listening while simultaneously worse in measurements.
Then there's always the argument of the classical fans that the 2nd harmonic emulates the concert hall reverb which is lost with close-miked performances. Tubes might simply be the perfect compliment to the weaknesses of the modern recording/mastering process, all by chance.
What do you and @tomchr think about the article? I found it a bit surprising and counterintuitive, but I'm not very well versed in these types of circuits.
Edit: In particular, I was surprised by the differences in noise floor.
For all the harmonics... are they in phase with the original signal? That'll change a lot of things.
I think non-linear distortion can get pretty complicated since it can be a function of fundamental frequency, levels, and even number of input frequencies and their relative levels and phases.
Knowing non-linear distortion is there, and how much as a function of a few input signals, is probably simpler than controlling the exact amount of non-linear distortion we may want out of a system for all input signals.
I'm not sure where I got these notions from, but these are the fiddly things in my head that are "painting the broad side of a barn with a bazooka"... so it's possible I made them up, but I feel like this is what I've read before. Again, super gross generalizations that potentially have no basis on reality:
- tubes have higher magnitude but lower order distortion
- solid state has lower magnitude but higher order distortion
- tubes distortion at its worse will pull things towards triangle waves*
- solid stage distortion at its worse will pull things towards square waves*
* I know what the simplified math for these are**, but throw phase crap in there and it goes to poopsville
** I lied, wikipedia knows what the math is
There you have it: the modern world in a nutshell.
I'm not aware of any physical effect in a vacuum tube that would introduce memory in the tube.
That's basically the argument I introduced a few posts back, which someone promptly shot down as BS. In my experience (and limitations thereof), some H2 (and maybe a tad of H3 and H4) will increase the sound stage. Whether you like this increase or not is up to you.
Separate names with a comma.