"Fast amps" - really fast or a bogus marketing claim?
- Thread starter caesar
- Start date
I can see the confusion. No, I only made a couple of distinct comments: 1) about SS implementations that I am familiar with (I have no idea what goes on in the tube-impl world); and 2) a basic statement about tubes themselves.
We refer to latency as 'smearing', there are probably many sources for it in a circuit, and one manifestation of it is in the squarewave graphs. I suggest you read the Soulution reviews in TAS, or Hammer's interview in issue 223, where they talk about their efforts to address negative-feedback-induced latency which affects the overall sound. I don't have the issues in front of me, by I vaguely recall them trying to drop this down to 2-5 picoseconds or thereabouts.
OTL as I noted can have that fast rise time because of the lack of Transformers.. That's a very fast rise time if the ms stands for microseconds which I believe it is ... Few SS aside from the Spectral, krell and Glodmund (some others?)of this world can claim this Speedy Gonzales speed... To repeat: aside from OTL, Audio Tubes amplifiers are not as fast on these objective terms as SS ... Even with OTL There remain the subject of settling time which is one aspect of "fast" and "speed". It is nice to rise but it is as important to stop when the signal is no longer present... and there the Capacitive nature of OTL outputs could be a problem. I don't know for sure .. Don Where are you when needed? MIA
????
Simple latency (constant group delay) would not cause "smearing" as all frequencies would be delayed equally. I find it hard to believe any audio amplifier on the planet would reduce latency to the 2 - 5 ps range. That would take an amplifier with much higher bandwidth (like, a microwave amplifier). Probably us, and that would be a fast audio amp.
A 100 W sine wave at 20 kHz into an 8-ohm load is about 40 V peak and requires a slew rate of about 5.1 V/us.
- P = V^2 / R, and assuming RMS then Vpeak = sqrt(2) * Vrms
- Slew rate = 2 * pi * f * Vpk
HTH - Don
A 100 W sine wave at 20 kHz into an 8-ohm load is about 40 V peak and requires a slew rate of about 5.1 V/us.
- P = V^2 / R, and assuming RMS then Vpeak = sqrt(2) * Vrms
- Slew rate = 2 * pi * f * Vpk
HTH - Don
Tim,
i'm not sure your comments were addressed to me or not; although i did write that phrase 'the speed inside the note' as a reference to my perceptions of the Found Music SET's based on both the 45 tube, and the 2a3 Tube.
as always; my comments have nothing to do with measurables; they have to do with my listening perceptions. i've owned other amps i consider to be 'fast' such as the Tenor 75 watt OTL tube mono blocks, and even more so the darTZeel solid state stereo amp and (my month with) the big dart mono blocks. up until i heard these micro-watt SET's the dart was my reference for fast in an amplifier.
'speed inside the note' is my attempt to try and describe what i am hearing with these two amplifiers. i have very strong solid references for how familiar recordings should sound. then with these amps i hear all this additional information revealed......this dramatic drop in distortion. it is resolving micro nuances in the signal by simply lowering distortion. all that comes at a cost of much lower power and less ability to do macro dynamics and space. so it's not a real world amp for most situations. but at this one thing it is unique in my experience.
of course; get the right speaker for the Found Music amps and they are real world amps.
i interpret less distortion as speed (faster). i see this in lots of gear. people accept as correct lots of gear that smears the musical message. until you hear it properly resolved, it stands as correct. then there is the realization that the distortion has been in the way.
that is simply my 'world view' based on my listening experience. YMMV, just my 2 cents, and all that stuff.
pardon if my descriptive phrase causes confusion.
Last edited:
In the case of Soulution that I brought up, the smearing is allegedly due to the delay of the loopback (negative feedback) itself; I think they explain it really well: the signal at the output at time X (now) is being fed back to the input, and while this loop completes, time has passed and the output is further ahead. Now, whether that claim is all true or not and to what extent, I'll leave it up to others to analyze. But *if* the waveform at the output is affected by any delay in the negative-feedback loop, then to me that would be smearing...
OTL as I noted can have that fast rise time because of the lack of Transformers.. That's a very fast rise time if the ms stands for microseconds which I believe it is ... Few SS aside from the Spectral, krell and Glodmund (some others?)of this world can claim this Speedy Gonzales speed... To repeat: aside from OTL, Audio Tubes amplifiers are not as fast on these objective terms as SS ... Even with OTL There remain the subject of settling time which is one aspect of "fast" and "speed". It is nice to rise but it is as important to stop when the signal is no longer present... and there the Capacitive nature of OTL outputs could be a problem. I don't know for sure .. Don Where are you when needed? MIA
No real argument against what you say. While my favorite amp the Moscode 402au is not OTL per se (mosfets). It does avoid the downfalls of output transfomers. It is fast.
Our friend Ethan et al says," there is nothing a tube can do that a transistor can't." I wonder if that cuts both ways.
But Gregadd the output is SS in the Moscode: MOSFETs. I do know that there are some characterisitics of Audio Reproduction Equipment that escapes a single parameter to fully describe them but speed doesn't seem to be one of them and in that sense SS are king let's add OTL if you please but the vast majority of Audio Amplifiers are not OTL ... So let's for once all agree that SS rules this universe. Let nos confuse the already difficult issues by describing some characteristics of reproduction as "speed" ... Going that route make everything murky and frankly difficult to grasp. Anyone who has heard a Spectral but also a few eletronics I find similar in their "speed": Burmester, Dar Tzeel, Edge, can attest that their attack and their decay is exceptional and to my ears, again aside from OTL (MOSFET Greg i-e SS) no regular (read transformers in the output) tubes can replicate this.. They may do for other people certain things better but for speed? ... Nahhh! Not tubes ... I have not heard the latest Krell ampps but didn't like the 2006 (?) EVo models I heard including the top of line. Nor have I heard the D'Agostino amps ... I must say I am in no great hurry to .. reports have been conflicting ...
i-e SS) no regular (read transformers in the output) tubes can replicate this.. They may do for other people certain things better but for speed? ... Nahhh! Not tubes ... I have not heard the latest Krell ampps but didn't like the 2006 (?) EVo models I heard including the top of line. Nor have I heard the D'Agostino amps ... I must say I am in no great hurry to .. reports have been conflicting ...Of course "speed or fast" are misnomers for reaction time. When it comes to tube the transformer(ss maybe) is the limiting factor. VTL/Manley make good ones.
Other factors contribute to reaction time such as power and damping factor.An under-damped or under-powered speaker will have several problems. These problems can slow the reaction time.
I wholeheartedly agree about Spectral. I am not really an authority on Burmeister, DarTzeel or Edge. they did fare well when I heard them at shows.
The D'Agostno is the real deal. I think the problem is that is no Krell clone.
If you've got the bucks to by an SS that matches or betters a tube amp. God bless you. I'll sidestep the ss/tube debate for now.
Other factors contribute to reaction time such as power and damping factor.An under-damped or under-powered speaker will have several problems. These problems can slow the reaction time.
I wholeheartedly agree about Spectral. I am not really an authority on Burmeister, DarTzeel or Edge. they did fare well when I heard them at shows.
The D'Agostno is the real deal. I think the problem is that is no Krell clone.
If you've got the bucks to by an SS that matches or betters a tube amp. God bless you. I'll sidestep the ss/tube debate for now.
I always thought speed referred to how fast the transistors can switch on and off and how everything before it in the circuit will allow it to do so. Some transistors are just faster than others.
As alluded earlier, negative feedback has been found to be an inexact solution to amplifier distortion. This is due to propagation delays (the very small but measureable amount of time it takes for a signal to move from the input of an amplifier to the output) which are a normal phenomenon of amplifiers. In order for negative feedback to work according to the math, it must be applied to counter the input signal in real time. Propagation delay in the amplifier circuit prevents this; the negative feedback will always be lagging the original input signal. This lag results in ringing effects and enhancement of the odd-ordered harmonics that the human ear/brain system uses to measure the loudness of a sound (specifically the 5th, 7th and 9th harmonics).
Atma-Sphere
And my understanding is that this is what Soulution attempts to address. I have yet to listen to their wares, though.
EDIT 6/26/12: Soulution claims a "few nanoseconds" delay in the loop, not picoseconds as I posted earlier... probably more believable
Last edited:
too much speed has issues though as well. over damped SS doesn't sound realistic to me---too much leading edge another way to put it and not enough of the whole note.
I have no idea what that means.
I said the whole note isn't present because of accentuated leading edge.
Sorry, musical terminology for playing slightly ahead of the beat. It can be a mistake, or a technique. Keith Richards is rather famous for it. Sounds like it might have a similar meaning to the effect you're talking about.
So you're saying an amp that's too fast can actually cause a part of the music/signal to play a bit too soon, accentuating the leading edge of the note? But the entire signal is not played too soon, (which would make it all arrive together, and early or late we wouldn't know the difference), just specific notes? And this phenomenon effects the rest of the note? So the note that arrives too early decays too soon too, ie: the whole note is not present? I've also got to assume, then, that the early arrival and early decay are not equal, because if they were, that would just balance out.
Interesting, and a new one to me. Is there a name for this kind of distortion?
Tim
