Zombie thread come to eat your brains!!
A lot has happened in the last 14 years...
|
Guidocorona, FWIW there is no audible difference between a damping factor of 60 and that of 1000... |
We did an evaluation of class D about 2 years ago. At that time the technology was still immature. My guess is that it is still; the best of the class D stuff, while not bad, does not hold up to the state of the art.
Yet.
It seems to me that tube and transistor technologies are both in a mature state, while class D has still a ways to go, and showing a lot of promise. If you know about price/performance curves, the significance of the preceding statement will not be lost on you! Something that you all want to keep in mind is that one of the bigger promises is that Class D costs about 1/10th that of conventional transistor amps, while the industry gets to charge about 1/2 as much at retail. It takes barely more than the ability to chew gum and walk at the same time to get why they are really here. We have 100 watt/ch. module here that is complete with heatsinks and connectors that fits inside a pack of cigarettes. It is average in its sound with respect to all the class D amps out there, but- it costs $25 in quantities of one... all you add is a chassis and power supply. The incentive to work with this stuff is powerful.
The same sort of thing happened with transistors back in the 50s and 60s- they take about 1/10th as much as tubes to make an amp of the same power, yet the industry gets to charge the same price. Of course the industry had to create a story about how the new technology was better than the prior art, just as was done with the CD. I invite you to consider that the story was made up for reasons other than performance, although as in many fields of human endeavor, performance can always be gained.
Class D lacks many of the artifacts of traditional transistor design, and are already good enough that any designer who ignores that fact does so at his own peril. It appears that class D has already engulfed the common consumer gear market- its ideal for iPods, boom boxes, car stereo and cheap receivers.
A simple tweek for many of the amplifiers is the power supply. Class D being what it is, the power supplies are subject to quite a bit of noise. A very simple means to deal with that very effectively is to use batteries- gel cells- charged by an overgrown wall-wart. Its amazing how easily that can be used to improve things. Like the CD, class D has a lowest common denominator quality. The modules are all fairly close in performance- the big differences are all about the power supplies and noise suppression- at least for now.
I'm really wondering where this can all go. For example, the modules right now are fairly complete units. But what would be possible if the input or even the output could be done with tubes? To understand where I am going with this, imagine an amplifier with a single conventional 300b that could make 50 watts and didn't make any heat. Tubes do not have speed/slewing problems- they can switch at very high speeds. Its the impedance matching issues that slow them down... I predict that in 10 years conventional transistor amps will be all but gone. |
Eldartford, I agree, unless you consider that commutation issues often limit semiconductor switching. A tube can switch at very high frequencies. Of course the idea is pie in the sky, but the point is that the technology is still unfolding and we really don't know yet where it will go.
Although the ICE module is one of the better sounding modules right now, the technology is changing so quickly that if you are 6 months behind, its already old hat. |
Muralman1, being one of those 'tube guys' I have to agree in part with you- this is something (at least for now) that I for one will not admit. IME, class D has a long way to go to beat what I see as the advantages to the amps that we make though I do not hold that out for my competition :) as our amplifiers are apparently unique even in the world of tubes.
I understand the cost, heat and size issues. However for the time being if the system is chosen based on those constraints rather than sound, then luck and luck only determines the sound of the system- no doubt why so many want to know which class D sounds the best!
If instead, the sound of the system is chosen entirely on the basis of sound quality, you get a different result altogether. Our amps were designed for performance/reliability as number one, all other aspects being secondary- we took a stand for quality in so many words. Class D was specifically intended to deal with heat/efficiency issues inside a cost model- consequently trying to overcome the issues that result flies in the face of its intention. Not to say this is wrong or bad- in fact, making the best of a situation has characterized a good bit of human behavior and consequent innovation. |
Muralman1 and Audioperv, I think the thing to look at is intention. If you start out with a given compromise, and then try to do the best you can, you wind up with something that is very good but with a compromise.
In the case of tube amps, compromise might start by choosing pentodes over triodes, as they make more power for their size. In fact if you look at transistors and the succeeding art, class D, that particular compromise of sound vs cost/size/weight/heat is one that as an industry we've been trying to beat for years.
So- what happens if you chose to do away with compromise? The intention is entirely different and so are the results. I agree that if you've heard a compromised (tube) amplifier, it might not best the class D amp that you have now, but the world of compromise has no meaning in the face of state of the art.
There is a common myth in audio that there is no 'best'. However, the English language tells us that it does exist- else the word 'best' would be meaningless. Usually we *believe* there is no 'best', since we have only heard that which is based in compromise. So we again come back to intention. |
So y'all understand, just because I make tube amps, that does not mean I'm blind to what is happening in the world of class D. I think I made that clear on my first post to this thread. Though Genesis did use a class D amp, it was a far cry from where the technology has gone today. Can you tell me what areas you think class D does not match your amp's performance? I think the area where we can show up a class D amp is in the area of soundstage presentation, the ability to be fast and detailed, yet relaxed in the highs, and last but not least bass impact. The latter seems to be one of the least understood, but most people also do not realize that a tube can be direct-coupled to a speaker. DC response in the output section, coupled with the ability to separate the driver and output section power supplies, means that the amplifier can operate at sub-audible frequencies without modulating its own power supplies, something that plagues many transistor designs. The issue here is: design according to the rules that the human ear uses, i.e. create circuitry that allows for the least offense to those rules. Audioperv, I'm not a nominalist, I am merely dealing with what is. As an example, I think we can all agree that our ears are the most important part of our listening, not our amps or speakers or taste of music. With no ears we would not buy audio products. So why in heaven's name ignore the rules that our ears use?? Yet many of the accepted design, test and measurement concepts do exactly that. Negative feedback is an example: yes, it lowers overall distortion, but- it *raises* the slight amounts of distortion that our ears actually care about. These are the high order odd harmonics- the 7th and 9th, which our ears use as loudness cues. Take a look at the distortion makeup of the typical transistor amplifier and the class D amplifier and the issue becomes clear. Without question this is one of the areas that defines the leading edge in semiconductor amplifier design. This is where the Pass 1st watt amplifiers gain their traction and where low or zero feedback tube amps get it too. Clearly this area challenges class D designers as well. Like it or not Class A still defines state of the art. But again, like I said earlier, any designer who ignores the advances and significance of class D does so at their own peril. |
Muralman1, here indeed is one the bigger issues of forum use in high end audio! We use a lot of the same language in description of system attributes, yet have no way of knowing if we actually mean the same thing with the same depth. So there is no way that I would know what you have, just as there is no way that you would know what I have without us hearing each other's setup.
The learning curve is without a doubt a serious concern. What I have seen of class D amps confirms your own findings- proper setup pays huge dividends (regardless of the technology). We deal with the same issues all the time with our amps too. I do wonder about some of the comments seen in the press that started this thread- how much was said based on setup issues? |
I am glad you are withdrawing from "it can be expressed in words, therefore it must be true" kind of logic. From my own perspective I have merely been dealing with what is. I notice audiophiles as hardened as I am, experiencing true rapturous epiphanies around the very same setup. Am I unable to grok true audioheaven? Or are those other audiophiles getting a little case of 'The Emperor's New Clothes"? Or should we perhaps accept somewhat soberly that we all may be seeking subtly differing things in our private quests for audio Nirvana? I think 'The Emperor's New Clothes' is a powerful issue in audio, perhaps because humans are more visually oriented than aurally. We get told that something is state-of-the-art (like a Yamaha receiver) and we pay the money and are so investing in the belief. But knowledge and belief are not the same thing despite how as humans we persist in trying to live our lives as if they *are* the same! So the result is a disparity, I think the result of one individual listening with their ears as opposed to another listening with the filter of 'status symbol'. How many times have we been sold 'new' as 'better'?? I am not saying that because class D is new that it is worse, nor am I giving it a whitewash: just from reviewing the comments of this thread, there is quite a range of experience. I think in the end that fact that an amp is 'class D' will not be enough for one to judge what it sounds like. Ultimately if an amp is sounding really right, it should sound like any other amp that is also sounding really right, shouldn't it? |
If a room sounds good at a show, I pay attention, but if it does not sound good, I don't let that deter me. I've had my own room sound lousy plenty of times... its really tricky to come into a room, set up a high end stereo and have it sounding convincing in a day or a day and a half (which is the timing that the shows give you).
If people keep coming back and wanting to hear more tracks I take that as a good thing, but I've never had a room sound as good as my system at home, nor have I heard *any* system at a show sound as good as what I have at home. I'm pretty sure I have a lot of company!
So we're back to the same deal- you have to audition this stuff. I've had the opportunity to do just that, and tweak the circuits as well. From that, and the feedback of customers, come my comments. My conclusion has been that while class D lacks many of the traditional solid state artifacts, that they do have artifacts of their own, but not ones that are common to class D in general (IOW very unlike traditional transistors in that regard; every transistor amp I've ever heard has identified itself as a transistor amp, class D does not identify itself as 'class D'), instead the artifacts seem to associate themselves with the specific product. That, by itself is a very good sign- it suggests that as the technology evolves, a truly neutral implementation could emerge.
Our auditions and customer input show that its not there yet, that we (as Atma-Sphere) have some breathing room for now. By no means has our own line of work been exhausted- the circuit is always suggesting new ways for it to be improved. So it will be interesting to see where things are in say, five years.
One thing that I have been intrigued with is that it is possible to build class D amps that lack the traditional style of feedback loop- a deadly sin as far as the human ear is concerned (and part of the 'traditional' transistor sound). That suggests that the amplifier can be designed around the rules of human hearing rather than the arbitrary standard of 'constant voltage' output. From what I have seen though, the open loop distortion is going to have to go down a bit yet... |
We have an internal design rule we call the 20-year rule that prevents us from using NOS tube types, IOW the tube type has to be in current manufacture. We've had this from the beginning, so its already served us well, as we have an aggressive update program for our older products.
Its a simple fact that semiconductors go obsolete and out of production at light speed relative to tubes. You can still buy new manufactured tubes that were designed 60 years ago, but I can think of quite a few semiconductors designed in the last 10 years that are already long gone.
The 'dwindling tube stock' idea is a red herring, unless a manufacturer has chosen to design around a tube that is no longer made, a foolhardy undertaking.
As a lot of the current class D amps are on monolithic modules using proprietary semiconductors, you can expect that if you need to service the amplifier in 10 years or the like that the device is going to be tricky to find. This is one sure way of knowing that the technology is still on its way up.
To get a better understanding of this, an example of a mature semiconductor technology is the lowly op-amp. You can still buy TLO82s brand new, despite their having been designed in the 1970s.
How many people on this thread think that they will have the same class D amplifier ten years from now- knowing full well that in far less time than that their amp will be superseded? |
Kijanki, FWIW your comment: My Icepower being smallest of the series has peak current 11A. Larger 201 has 20A and largest 501 has about 50A. has an ambiguity. The 50A peak current you mention- is that the output? Assuming a 1 ohm load and 50Amps, that's 2500 Watts. Assuming a 2 ohm load that's 5000 watts... So into what impedance does this spec occur or is it 'marketing'? |
Kijanki, I agree completely- and will add that many of today's specifications exist to sell certain types of amplifiers rather than get to the truth of the sonics. |
Kijanki, not to detract from your comments of the review of the review, but just so you know the numbers that are quoted are not speaker current. If it were:
First, reality:
For 100w into 8 ohms, the current would be about 3.54 Amps.
For 100w into 4 ohms I = 5Amps, for 2 ohms I = 7.1Amps
then the 'claim':
working it backwards, 50A into 4 ohms is 10,000 watts
into 2 ohms its 5000 watts.
So- 20Amps into 4 ohms is 1600 watts. But your amp does not make that kind of power!
I agree these numbers are common in a lot of amplifiers, the 'FWIW' part is that the numbers are *not* current in the speaker. I suspect they are power supply numbers, that number being the current output of the supply when shorted for 10ms. Our MA-2 amplifier which is an all-tube amplifier measures about 80 Amps using this technique... |
Guidocorona, I was merely commenting on a spec that clearly was not possible as it was represented. I also demonstrated that with the math.
If you think I was taking a pot shot, I apologize. However, let me point you to a paper:
http://www.atma-sphere.com/papers/myth.html
The paper is about 5-6 years old, written long before class D was commonplace. The fact of the matter is that amplifier manufacturers have been placing specs like this before the public for years. I would prefer that the spec be placed in its meaningful context: that it is not speaker current but something else.
I've been watching the class D technology for some years as you must know if you've been reading through this thread (which has grown quite quickly!). Is it a threat to what we do? I don't think so, not because it might be better or worse, but for the simple reason that I am not going to be so close-minded as to think that nothing can beat what we do: there is no integrity in that.
We've been dabbling with class D for about 4 years and I plan to continue. As I've mentioned before, its a newer technology, and obeys price/performance curves that all new technologies follow. Right now we are in the middle of the curve. What that tells me is that we don't know yet where the technology levels off as it matures: it could go a long way yet! I apologize for my pragmatic outlook, but it is what has kept me in business for the last 30 years... |
Unfortunately companies are racing for best specifications. I would advise to take specifications, promtly discart[sic] them and just listen. Still the best advice anyone can give. |
For my ears, Class D amp builders are not trying to achieve a sound like tubes or solid state. They are trying to amplify the truest signal to their ability. This comment sounds like a strawman to me: I know a fair nubmer of conventional amplifier designers (tube and solid state) who have the same goal- this is nothing unique to class D! |
Mrtennis, by any chance do you have a recording made in the space that you just heard? My experience has been that its apples to oranges comparing live to recorded unless you are familiar with the space that the recording was made in.
I've done a lot of local recording with nice microphones and a portable tube recorder. Some of those recordings have made it to LP and CD. If I hear them played back properly I can always hear the hall- they easily remind me of where they were made. I've come to the conclusion that this is a major part of having a 'reference'. |
Muralman1, I appreciate that you like what you hear. My point is that to really know if it is right, it is very useful to do on location recordings with professional recording gear (and with the intention to do a good job...).
The master tapes then are very useful in letting you know how a system really sounds, because you were there at the recording. You know what the hall sounds like, where the instruments were placed, how far the mics were from the musicians, that sort of thing. This is assuming that you have good recording technique and the equipment is up to snuff... |
Muralman1, I am sorry you took offense at my comment, but if you look at my last post and your initial reaction, it will be obvious that you attached meaning where none existed.
My point was that if you go ahead and make good quality master tapes, you will be in a position to use those recordings as reference tools, for all the reasons that I outlined earlier.
Will the results mean something? I have no idea. All I know is that these tools were invaluable for me personally and others who have done the same thing.
IOW, a system may sound great, but its always in respect to what. When you have master tapes, now you have an absolute. |
Muralman1, you're in a good position to record that viola- and then play it back. *Thats* what I'm talking about. You already *know* what the instrument sounds like.
That's the technique I use when I am auditioning anything- speaker, a tweak in one of our products, a class D ampilfier- whatever. Sure, I play a variety of recordings, but its always useful to be able to play one where you were at the recording session.
To address an earlier comment- the distortion that a tube makes is not something that is cast in concrete. I have found that quite a bit rests on design consideration as well as materials. Not all tube amps have a predominant 2nd order harmonic- that's really a character of SETs. Push-pull amplifiers have even-order cancellation.
If you look at the distortion makeup of a lot of the class D modules, the distortion can be quite high for a 'transistor' amplifier. We have a module here that is 10% at full power. We've not measured the spectra, but I can tell you that the distortion is under 0.5% or so to about 50% of full power. The IM is actually higher than measured on our own production amps. Of the two, IM is usually more audible than THD. The appearance is very much (in the case of this module) that the ear is ignoring a lot of the THD, i.e. it appears to be even-ordered. IOW it seems to have a lot in common with SET amplifiers! I thought you might find that interesting...
By no means is the module that we have (Phillips) representative of the entire field. It would be interesting to see if there is a correlation either negative or positive with what we hear in these amplifiers as opposed to what is measured! A negative correlation, FWIW would put the behavior in the same realm as tubes, where much of the distortions made are ignored by the human ear. |
Muralman1, I've not checked out all the modules that are out there, and I was careful to comment to that effect.
FWIW though, distortion has nothing to do with what impedances an amplifier will drive. The two issues are unrelated. Neither is THD a good indicator of how an amplifier will sound. The current test and measurement paradigm does not generate specifications that are intended to let us know how an amplifier might perform sonically. |
Guidocorona, you may not believe this but I agree with you- I don't like 'tube' colorations either. I think many of those colorations have to do with design considerations rather than the tubes themselves.
Consider that because you have heard some tube amps which have coloration does not mean that all of them are like that.
I don't really like *any* colorations (tube or solid state). Spending a lot of time in the studio, hearing the live feed as opposed to the recording, how different mixers, speakers and amps behave will do that to you.
This is one of the intrigues of class D for me as they often lack some of the more heinous colorations that I associate with transistors and tubes. For all that, my experience is that class D amps often have their own unique sonic fingerprint that varies from brand to brand as it does in the scale of execution. Its getting control of those variables that seems to be at the cutting edge of this art. |
