5 watts is plenty. Attach that to a good quality pair of 100dB speakers, and you get 106dB peaks. Plenty of dynamic room to render the operatic voice correctly ...
FRank
Power . How much do we need...
- Thread starter FrantzM
- Start date
FRank
Ahh, but nothing to stop you running the tubes in parallel, we're talking limitless money, or stupidityperhaps, how about 100 of the 2a3 tubes in such a configuration: be able to hold its own against a Krell!
Frank
the whole idea of an SET is simplicity. your approach might result in more power, but you would lose the low distortion advantage to the tube.
the reason there are so many different approaches to SET and even tube amplification is to optimize the purity while providing more power. at one end of the scale you have ultimate purity, at the other end lots of power. there are low distortion SET's which are hot rodded to get more power, such as my 3 watt 2a3, which as 2 to 3 times the normal power, as well as stability into a load, of a normal 2a3 amp. then there are the best of the big power tube amps which have relatively low distortion. but the 3 watt 2a3's don't get as close to the power of a 300+ watt big tube amp as the best of those get close to the low distortion of the 2a3. but the 2a3 still has a pretty significant advantage in low distortion. but you do choose your priority.
just because a big high power amp is not as low distortion as the best low distortion SET does not mean that the high power amp is not good enough. that's a matter of opinion. it's only when you say that this high powered amp is better at everything than the low powered SET where there is a conflict. lots of output devices and circuitry do have their effect on signal purity.
How loud does one listen to music and how far away from the speakers is one's listening position? The math is the math; it is not subject to dispute.
Regarding distortion of flea or, for that matter, high powered amps, there comes a point where the ears fall short of actual measurements. As such, to proclaim a particular amp has the lowest distortion, lowest noise floor, etc., one needs reliable, falsifiable data. This is not something that is determined by one's personal taste or ears, since we're all different. Show me the data.
Regarding distortion of flea or, for that matter, high powered amps, there comes a point where the ears fall short of actual measurements. As such, to proclaim a particular amp has the lowest distortion, lowest noise floor, etc., one needs reliable, falsifiable data. This is not something that is determined by one's personal taste or ears, since we're all different. Show me the data.
Mike
Let me try to answer your post . Allow me some tangents.
Back in the early days of WBF, Caesar (Him of several very intelligent questions) asked about Education and Audiophile. By Education I think he meant a learning process. We do learn to evaluate equipment and there is some consensus among audiophile of what is good sounding. Most of us will never declare any Bose great sounding. Most people can recognize if something sound very close or similar to a real instruments. It takes some training to grasp nuances and to develop or try to develop a vocabulary that convey auditory sensations or the closeness to the real thing of a given reproduction …
“Good Sounding” is one of those terms that are so fraught with subjectivity as to be , for me at least difficult to use in this context. What sounds good to you might sounds good to me or might not … I do however know that it is entirely impossible to give a good simulacra of a soprano going full tilt in your room with 5 watts on your speakers.. Even Less a Wagnerian score or a Mahler symphony, Do not approach it with Shostakovitch 10th or Kanye West .. Ok you get my drift …
I am not of the advice that it is that difficult to build a powerful amplifier. I would surmise to you that The mighty Top of the line Dar Tzeel which you heard at your place is a supremely great sounding amplifier and on your and likely most speakers, will present a very believable simulacra of a soprano going full tilt or the aforementioned symphonic work or a Jay-Z score… There are numerous amplifiers both tubes and SS that would do the same .. ARC, VTL, MBL, Burmester, Gryphon, Boulder, ML, Threashold, Bryston, etc.. You may not like them and that is OK .. yet when it come to similitude with the real thing.. They would be at ease ... To me that what power is all about and to me that counts.
One may prefer certain thing that a low power amplifier does on a medium sensitivity speaker and that is fine. It is likely limited to a small genre or number of musics or music pieces… I also am surmising the that the flea-power amps highlight some part of the frequency band because their response in other parts are not as good … kind of reducing the bass volume (assuming that a speaker can do that) to listen to more mid-range.. If at the end one derives pleasure from this… Fine. Mission accomplished … But In term of verisimilitude, that is like-ness to the original even , I am not sure the flea-power will match the Dar Tzeel on most if any music score … I may prefer what it does and often what one prefers may not be what is accurate .. Kind of preferring Tang to Orange Juice an example that many people don't like but which to me conveys the idea very well. If you like something and derives pleasure from it .. That is to me fine ... End of debate..
And there is the pure objective aspect of the story when the flew-powered amps find itself out of steam what does it do .. What can it really do? However superlative its performance was up to this point which is reached very quickly with most modern speakers? It clips and while clipping is what guitar players wish .. Reproducing a wave with any semblance of accuracy is not the province of a clipping amplifier.
Thus my position that all things being equal … The more power I can get to my speakers, the more I will provide them… And frankly for the price I see some of these flea-powered amps going for I would say that I will be more comfortable with gob of watts than with sub-10 watters … For the speakers I am thinking about these days (yours being very much part of my list maybe at the very top of it)
Frantz,
i agree with the perspective of your post. and my system development also agrees. i have built a room and put together a system which has no practical dynamic limitations. yes; i recognize that the big dart monos would bring an extra level of ease and grip to my mids and tweeters, but my stereo dart truely delivers these things already. i can easily live without the difference between the 2 darts....or a better way to put it might be that i prefer the stereo dart to any higher power amp i can afford that i am familiar with.
my reaction to your post was to the lack of qualifications. either more power is always better, or it's not always better. or, it's not that simple.
i think it's not that simple, and more power is not always better since many times more power involves unacceptable compromises.
my flirtation with flea powered SET's has not changed my basic perspective on what i think a system should do. OTOH it has showed me what low distortion truly is. my hope is that i can somehow have that view as an option from time to time.
I always thought Martin Collom's approach when reviewing speakers was very sensible, as a means of assessing and comparing speakers. Measure the real world sensitivity, take the recommended maximum amplifier power, and work out the "Approximate maximum sound level (pair at 2m)". So for a (quite old) B&W bookshelf unit: 89.5dB/W, 100W amp max., works out to 105dBA peak volume.
Frank
MIke
It is alwways apleasure to discuss with you ... It will be a pleasure to share long listening sessions with you ( see forthcoming PM) ...
I like the notion of simplicity .. A triode is on the surface a very simple design.. The simplest form of tubes ... In my view a transistor is also very simple .. three pieces of rock bonded together ... For these "simple" components to work, many other "things" must be added"Resistors, Capacitors, sometimes inductance.. All, elements that make the resulting system a little less .. simple .. more complex ... The presence of transformer make many tubes more complicated than one would think introducing a whole lot of different problems unless one goes OTL which comes with its own set of caveats ...
As things go .. From what I tend to go with being an engineer namely reproducible, repeatable data.. a SET is not the lowest distortion topology .. Measurements, I know it is a dirty word, tend to go against SET.. Let me grant you that they can sound good, downright beguiling, I have heard the Lamm and Jadis and VTL SET and they are magical...
I also understand the idea of having it both ways .. I believe someone else here has done the same (Tomelex?) ...
My main point is that at this point, there are several high power amplifier that would provide superlative performance that leaves nothing to be desired. Performance that would rival everything these fle-powered amps do ... I also understand the point that given the choice between great sound and lot of power I will go toward great sound .. I did take the Burmester 911 over a 650 watts behemoth that would have driven my MG 20.1 extremely well .. The thing remains that the 911 is not a puny amp: it'll drive incredibly difficult speakers to ear-splitting level and be delicate the next second ... I have never heard it clip on the 20.1 for example and I have witness it driving the difficult MBL 101 at a friend place when the spectral DMA-180 was driven to its knees ... And it remains one of the best amps I have ever heard ...
@Orb forthcoming reply to your most interesting question
It is alwways apleasure to discuss with you ... It will be a pleasure to share long listening sessions with you ( see forthcoming PM) ...
I like the notion of simplicity .. A triode is on the surface a very simple design.. The simplest form of tubes ... In my view a transistor is also very simple .. three pieces of rock bonded together ... For these "simple" components to work, many other "things" must be added"Resistors, Capacitors, sometimes inductance.. All, elements that make the resulting system a little less .. simple .. more complex ... The presence of transformer make many tubes more complicated than one would think introducing a whole lot of different problems unless one goes OTL which comes with its own set of caveats ...
As things go .. From what I tend to go with being an engineer namely reproducible, repeatable data.. a SET is not the lowest distortion topology .. Measurements, I know it is a dirty word, tend to go against SET.. Let me grant you that they can sound good, downright beguiling, I have heard the Lamm and Jadis and VTL SET and they are magical...
I also understand the idea of having it both ways .. I believe someone else here has done the same (Tomelex?) ...
My main point is that at this point, there are several high power amplifier that would provide superlative performance that leaves nothing to be desired. Performance that would rival everything these fle-powered amps do ... I also understand the point that given the choice between great sound and lot of power I will go toward great sound .. I did take the Burmester 911 over a 650 watts behemoth that would have driven my MG 20.1 extremely well .. The thing remains that the 911 is not a puny amp: it'll drive incredibly difficult speakers to ear-splitting level and be delicate the next second ... I have never heard it clip on the 20.1 for example and I have witness it driving the difficult MBL 101 at a friend place when the spectral DMA-180 was driven to its knees ... And it remains one of the best amps I have ever heard ...
@Orb forthcoming reply to your most interesting question
Last edited:
My understanding is that it was always per the manufacturer's rating. These days I note this is not given so much, more like the minimum power only for satisfactory performance is typically nominated by the maker. In part I think this is because most speakers can take transient high power peaks without damage, so that B&W would actually have gone louder if done with due care. Woofers can bottom if driven too hard, but midrange and tweeters will cook is stressed overly, I have never noted maximum excursion as being a limiting factor for the latter.
Frank
I mentioned average SPL because I am not able to measure peak (I do not have the equipment). But I guarantee that my amp was clipping (or whatever you want to call it) at a level that most people would not describe as 'loud'.
As for the sweeping statement about Sander's post:a) did you read it? and , b) what did not make sense?
This:
If what he says here were really true the music would sound a distorted mess, if a level of clipping that was so obvious on a 'scope occurred. But then again he could be talking about a remastered Iggy Pop album :b.
Then he continues, as if this were typical:
If your speakers were really handling 200W of clipped waveform then your eardrums would be shredded, and there would be smoke pouring out from the tweeters, an extremely unrealistic situation for a normal system.
Finally:
This implies that Martin Colloms doesn't have a clue what he's talking about in his reviews: that the amp has to be capable of driving that speaker to close to a 120dB peak at 1m to not be clipping ...
Frank
Well I did precisely what he suggested with the oscilloscope on a piece of choral music (amongst others) and saw precisely what he describes. At no point did the music sound like a distorted mess. I felt that my amp was "running out of steam" which is why I tried this experiment and now I know why.
The clipping is very quick and not there all the time but it is very obvious and easy to see.
Simple test, give it a shot
Although coming back to clipping tweeters,
it is worth noting that it seems many (maybe most) will have -20db energy from 2khz onwards when compared to the lower frequencies.
I am basing this on the reviews in HifiNews for their hi-rez where for several months now they analyse the albums and show peak-rms amplitude against frequency (with time applied in graph), in the magazine they show one of the tracks and I have to say consistently anything about 2khz is roughly -20db minimum compared to say 500hz and lower.
Cheers
Orb
it is worth noting that it seems many (maybe most) will have -20db energy from 2khz onwards when compared to the lower frequencies.
I am basing this on the reviews in HifiNews for their hi-rez where for several months now they analyse the albums and show peak-rms amplitude against frequency (with time applied in graph), in the magazine they show one of the tracks and I have to say consistently anything about 2khz is roughly -20db minimum compared to say 500hz and lower.
Cheers
Orb
Okay, this is becoming interesting; I suspect that you had fairly severe voltage sag. Did you get an approximate measurement of what the voltage peak was when this "running out of steam" occurred? And do you know what the nominal voltage rails of the amp are?
The interesting thing is that if you do a simulation of a power supply with real, rather than perfect components, it's very easy to make the power supply collapse. So if you can do this with a computer model imagine how a real setup will perform!
Some time ago Tom Danley mentioned on this forum seeing this sort of behaviour with a well regarded, high end amp; he contrasted its "bad" behaviour with pro amps, which were fine ...
Frank
But that is of the audio track when the power supply is NOT clipping! This is why a cheap, low powered amp can destroy the tweeters of an expensive, high power speaker: the waveform of a clipping signal has a huge boost in the level of high frequency energies, courtesy of the sharp corners of the clipping, which get fed straight through to the tweeter. The kids have a loud party: bye, bye, little speaker!
So, yes, there is a definite danger to these drivers if you constantly feed a highly compressed -- translation, voltage sag leading to possible early clipping -- signal at high volume for lengthy periods to the speakers ...
Frank
Mike, I could come up with some obscure product you'd never seen, or some bizarre tweak of my own invention you were unable to try, and claim that it renders the best sound - capitalized, italicized and underlined - then tell you it is not my opinion, but a fact, and if only you could hear it, you would agree. But it's a child's argument. It's not worthy of decent debate, or even on-line audiophile discussion. The bold humility of accepting that others may not hear things the way you do -- IMHO -- the understanding that not all people take your path, and theirs is not necessarily wrong -- YMMV -- it really shouldn't be that hard to embrace and accept. And you're talking about distortion in tubes, so you're ignoring or denying all the available data to get to your point of view, making it a purely subjective one in the process.
The bottom line: Is this "reality" perception or fact? If it is perception, it is unquestionably your opinion and my mileage may (and probably will) vary. If it is a fact, which is heavily implied by your remarks about distortion, these are facts that are measurable and repeatable. But let's start with just once: Point me to the distortion figures of this amp.
Tim
PS: i
Given that "many times more power" than SET is as common in high end audio as opinions, it should be really easy to elaborate on these unacceptable compromises.
Last edited:
No I didn't get a measurement and nor do I even know what rails are in any meaningful sense.
The interesting thing to me is not what you can get a simulation to do, but rather the fact that a 50w amp can clip into 97dB speakers at normal listening levels.
And the reason is that at the transient moment when it's clipping it's no longer a 50W amp, it's more like a 10 or even 5W unit. Huhh?? Well, if the voltage rails are sagging very badly, which is what those scope results indicate, then the actual power delivery at that moment is much, much lower than it nominally is, what the spec's say. If the rails drop to half their usual voltage; knowing power delivery goes by the square of the voltage, power into load has dropped to a quarter of rated!
That's why I queried the voltages involved, would tell you a lot ...
Frank
Just came across this, http://www.bottlehead.com/store.php?crn=44&rn=447&action=show_detail, after looking around about 2A3/45 amps, and the comments by lmalinofsky were very edifying. He of course is getting genuine over 100dB sound levels through using Klipschorns and, probably like Mike, is getting the "real deal" of what "flea" power can do. The language used to describe the sound is exactly that which is used by people who "get it", nice to see the circle is widening ...
Frank
Frank
Yeah.
But the point is clipping is much more likely to occur for lower frequencies as the 2khz and above are another -20dbfs when looking at it from real music-recordings.
Challenge is that music is not same as a single static-predertimined sinewave.
So to clip the higher frequency (which would then make it through the crossover to the tweeter) you would need to push the amps waaay beyond its ability with lower frequencies.
Just my take on it anyway.
Cheers
Orb
Hi
This will not end the debate about preference and subjectivity and the unreliability of many of our observations. The acuity of our perceptions/sense vary from moment to moment and is very dependent of psychological factors. Most of people on this board or other audio boards know this for a fact. We know how easily our eyes and snse of smell, taste and touch can be and are regularly fooled . SOme of the things we enjoy depends on our eyes being fooled by a succession of still images that the eyes and brain nicely merge as a fluid moving image...
We are regularly fooled by our ears and that is where the audiophiles, myself included cross the line . It's all good, it causes no damage, has not (yet ) provoked any wars aside from the heated audio debates that we so enjoy . I have seen all kind of argumentation artifices used in audio discussions to find a way to rationalize differences that in many cases, not all I hasten to add, are not audible ... I am by no means saying that we know it all and that everything has been studied, there are surprises waiting around many corners and we better open our mind to the fact that some difference not yet fully measured or understood may exist.
There are however clear cases where it is more our bias expectations that anything else at play .... and the power required to drive a given speaker is one of them ... Most speakers are grossly inefficient. The sensitivity numbers do not do justice to the vast inefficiency of our loudspeakers. They are not even 10% efficient ... for One (1) electric watt furnished to most speakers we don't get much more than 0.01 watt of acoustic power radiated at 1 meter ... That would be for a 100 dB/1W/1m speaker a 90 dB speaker has a 0.1% percent efficiency .. for one electric watt it only produces 0.001 acoustic watt /m2 ...
Most modern speakers have a network called a crossover. To use a current term, a router for frequency bands and this thing is not necessarily efficient in itself and it behavior in term of load presented to the speaker is not linear, not constant with frequency and even with the power furnished ... A lot of the watts pushed to the speaker are lost .. there and elsewhere within the speaker that would at tend only push a little trickle of this power in acoustic power ...
And we have been making great powerful amplifiers for a while ... and they sound good .. And many measure very well and many compensate for the inefficiency of our speakers of any nature by being ....well ...very powerful and very close to a voltage source, i-e maintaining the voltage across their terminal very stable regardless of the load .. Most of these if not all are not flea-powered and few if any are SET ...
I will leave to anyone to draw their own conclusions but the above seems to me a case for more power not less...
P.S. If a person is using a 100 dB speaker he/she can use a 3 watts amp and have plenty of SPL... Few of the speakers that audiophiles call the "best around there", are close to these numbers.. Very few with the most highly touted mostly hovering around 95 and more often less, and a few squarely under 90 db pushing hard toward the mid 80's an efficiency of 0.0001% . That's inefficient
This will not end the debate about preference and subjectivity and the unreliability of many of our observations. The acuity of our perceptions/sense vary from moment to moment and is very dependent of psychological factors. Most of people on this board or other audio boards know this for a fact. We know how easily our eyes and snse of smell, taste and touch can be and are regularly fooled . SOme of the things we enjoy depends on our eyes being fooled by a succession of still images that the eyes and brain nicely merge as a fluid moving image...
We are regularly fooled by our ears and that is where the audiophiles, myself included cross the line
. It's all good, it causes no damage, has not (yet ) provoked any wars aside from the heated audio debates that we so enjoy . I have seen all kind of argumentation artifices used in audio discussions to find a way to rationalize differences that in many cases, not all I hasten to add, are not audible ... I am by no means saying that we know it all and that everything has been studied, there are surprises waiting around many corners and we better open our mind to the fact that some difference not yet fully measured or understood may exist. There are however clear cases where it is more our bias expectations that anything else at play .... and the power required to drive a given speaker is one of them ... Most speakers are grossly inefficient. The sensitivity numbers do not do justice to the vast inefficiency of our loudspeakers. They are not even 10% efficient ... for One (1) electric watt furnished to most speakers we don't get much more than 0.01 watt of acoustic power radiated at 1 meter ... That would be for a 100 dB/1W/1m speaker a 90 dB speaker has a 0.1% percent efficiency .. for one electric watt it only produces 0.001 acoustic watt /m2 ...
Most modern speakers have a network called a crossover. To use a current term, a router for frequency bands and this thing is not necessarily efficient in itself and it behavior in term of load presented to the speaker is not linear, not constant with frequency and even with the power furnished ... A lot of the watts pushed to the speaker are lost .. there and elsewhere within the speaker that would at tend only push a little trickle of this power in acoustic power ...
And we have been making great powerful amplifiers for a while ... and they sound good .. And many measure very well and many compensate for the inefficiency of our speakers of any nature by being ....well ...very powerful and very close to a voltage source, i-e maintaining the voltage across their terminal very stable regardless of the load .. Most of these if not all are not flea-powered and few if any are SET ...
I will leave to anyone to draw their own conclusions but the above seems to me a case for more power not less...
P.S. If a person is using a 100 dB speaker he/she can use a 3 watts amp and have plenty of SPL... Few of the speakers that audiophiles call the "best around there", are close to these numbers.. Very few with the most highly touted mostly hovering around 95 and more often less, and a few squarely under 90 db pushing hard toward the mid 80's an efficiency of 0.0001% . That's inefficient
Similar threads
