Colin Posted January 16, 2005 Share Posted January 16, 2005 Defenders of solid-state equipment are quick to say that a QUALITY solid-state amplifier can sound just as good on big ole horns as tube equipment. And that much does seem to be true. I have heard the incredible Pass Laboratories Balanced Single-Ended X250 monster amplifier (which doubles, then DOUBLES again, output as resistance goes higher) with super-sensitive walnut-oiled Cornwall 1s, with B2 crossovers. The Pass cost about $4Kused. I hear my 70s vintage harmon/kardon solid-state single-transformer 330B receiver on my classic Klipsch corner Khorns. This slim receiver cost $5 in a yard saleand was gift. And it is true. Both of these solid-state amplifiers do an incredibly good job with super-sensitive big ole horns. In fact, they both sound just as good as my flea-powered tube Bottlehead 2A3 Paramour amplifiers. The Pass of course had unlimited energy during the loudest movie explosions, the harmon/kardon is a little more solid in the mid-bass, the Paramours still retain as edge of quickness and sweetness in the mid-range, while imaging a little better too. I dont have access to the Pass anymore, my friend was called up for National Guard duty, but the other two have not required an ounce of maintenance or care in the last few years. I know why the Pass might be so superior. It has oodles of power and high quality components in a SET Class A design. I still dont know why the harmon/kardon is so good, or why nobody is making a receiver with this design anymore. I understand how the Class A SET design and the 2A3 tubes of the Paramour leads to its sweet sound. But if I was looking at their specs, how could I tell which one is going to sound the way that they do? How does one judge an amplifier, or any other component from its specs? Unless you can see all the charts and measurements that Stereophile provides (and even then, I need plain English language interpretation), what clues do the specs give neophytes? How can I judge an amplifier by its specs? Quote Link to comment Share on other sites More sharing options...
Klipschguy Posted January 16, 2005 Share Posted January 16, 2005 Colin, As you already know, it really comes down to dimishing returns: Speakers & room >>>>>>>> amplfier. Think about it. Great speakers in a great room will sound great with just about any reasonably well designed receiver with "good specs." Conversely, lousy speakers in a lousy room will sound bad even with great amplification. Interstingly, great speakers in a lousy room cannot reach their potential even with superb amplification (my Cornwalls sounded much better in my former house, although they still sound pretty good - time for a new house?). IMHO, speakers and the room are at least ten times more important than the amplifier (as long as the amp is reasonably clean). Thus we see the dimishing returns of the amplifier link. (Not to mention that amplifier technology is so good and relatively cheap that we see dimishing returns from one amp to another.) No, I'm not saying all amplfiers sound the same, or that an expensive amp is not worthwhile - they just do not carry the same ability to affect the overall sound of the system like the speakers and room do. Please note: it is important for EVERY link to be as strong as possible for the best overall sound - plus its fun! Just my experience and opinion. No flame wars, please. Warm regards, Andy Quote Link to comment Share on other sites More sharing options...
Klipsch Employees Trey Cannon Posted January 16, 2005 Klipsch Employees Share Posted January 16, 2005 In two words.....you cannot. Almost never will you find amps / receivers with the specks given the same way. There is no "law" as to how you rate an amp. However, it is not as bad a in the car audio world. Demo, Demo, Demo..... The only way to know. just mho. Quote Link to comment Share on other sites More sharing options...
Duke Spinner Posted January 16, 2005 Share Posted January 16, 2005 spec's .. like headroom, and class of operation may point the way ... demo, demo, demo ...... i spent most of my money on ........ 1) speakers .. ( lotsa them ... ..) 2) pre-amp .. perhaps a tube in there 3) source, ..be it a turntable, or Cd i run Crown com-techs for power amps, as i find this is the least critical component of MY system Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted January 16, 2005 Share Posted January 16, 2005 One method of optimally matching any source to any load is in examining their Nyquist plots (Oft referred to as Heyser spirals). These plots of the frequency dependant complex impedance provide ALOT of amazing insight in to the performance of the said units, and the coser the match, the better the two units will perform. After all, you do not want to match an amplifier that is capacitive at the lower several octaves with a subwoofer, or conversely, an amp that is inductive at the MH frequencies with a MH driver. You are simply mismatching the source and the load and expecting the amp to do what it does NOT do well! The sound will suffer! It is amazing how audible the correlation of subjective listening and specs are when this tool is used to match components. But that being said, I will maintain that perhaps the single most significant interaction is that of the tranducers within the the acoustical space of the room itself in the time domain. And this simply cannot be dealt with effectively with an RTA and an EQ! This must be dealt with first in the time domain, and once the early reflections, etc. are addressed and the system brought into a miniumum phase condition (all arrival times being coincident), THEN and only then can you effectively use an EQ. Until then, all you are doing is shifting the phase of each signal via the L&C components of the filter and modifying the resultant polar and comb filtering standing waves resulting from the superposition of the various acoustical sources (including reflections). In other words, as you change the phases, the polars and standing waves/nulls tend to move slightly. You don't fix the problem, you simply move the apparent problems around - within limits. Various tools can be employed to ascertain the room and component responses, the most complete being the TEF20 analyzer. This will generate the Nyquist plots, apparent acoustical centers for each driver (useful for crossover alignment in the time domain), arrival times and reflected energy, paths, and using the reverse resolution function, identifying the relctive 'spots' and allowing proper treatment of them in quick order utilizing the laser POGO system. A marvelous tool! Quote Link to comment Share on other sites More sharing options...
Colin Posted January 16, 2005 Author Share Posted January 16, 2005 "capacitive at the lower several octaves with a subwoofer, or conversely, an amp that is inductive at the MH frequencies with a MH driver" how do you read that from the Heyser spral? Quote Link to comment Share on other sites More sharing options...
MrMcGoo Posted January 16, 2005 Share Posted January 16, 2005 I agree that specs alone are not going to provide reliable answers about amplifiers. Part of the decision as to what a "good" amplifier is personal and subjective. Solid state pure class A amps tend to be very heavy and run very hot. Class A SS amps need to be warmed up before they sound good. Such amps are impractical for multi-channel sound IMO, but they can sound quite good. Note that all of the required information is readily available for me to avoid these class A amps, provided I realize they do not fit my lifestyle. What do I look for in an amp before I listen? I look at a solid state amps ability to drive low impedances down to 2 ohms or 1 ohm. The ability to follow the impedance curve means that the amp should have better frequency response and do well on dynamic peaks. Does the amp have a reputation for sounding good on high sensitivity speakers? In a multi-channel setup, does the amp have enough power to run all of its channels simultaneously when dynamic peaks hit? No 7 channel amp that I am aware of can do this, so I use a five channel amp for the center and surrounds and a stereo amp for my mains. Finally, I need to supply my amps with enough power to run properly. I have a 20 amp circuit for my theater, but two 20 amp circuits would be better. Since I don't have them, amps that do not waste power are usefull. Bill Quote Link to comment Share on other sites More sharing options...
Klewless Posted January 16, 2005 Share Posted January 16, 2005 Here's my take on measuring amplifier quality. Actually these comments apply to any and all of the components. 1. Have enough power to handle your desired sound levels. 2. By the number of items you can hear and count within the source. 3. How close to your conception of real the "system" has presented. 3A. You cannot singly hear any one component without all the others being in the chain. 4. Again how close you perceive the overall sound stage to sound real. I know this was not a technical response. Nobody has determined the environment with which to evaluate the audio chain. For me the evaluation stops when I find myself listening to the music rather than the system. This then presents the age old question of "How HI the FI? Quote Link to comment Share on other sites More sharing options...
Guest Anonymous Posted January 16, 2005 Share Posted January 16, 2005 i agree with trey demo demo demo Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted January 16, 2005 Share Posted January 16, 2005 Oh, how to interpret the Nyquist plotYou would ask that! ;-) First, lets define the graph that it is displayed upon in 2-space, and then we will enlarge it to 3 space. And then, if you ever wondered what in the world they were talking about for all those years of math and circuits classes you may have taken, the concept of real and imaginary numbers will FINALLY make sense! The graph has two axes. The x axis is the real component. The y axis is the imaginary component. And the z axis (which we will get to in a minute) is the z axis. What you see if you are looking a 2 space representation of the Nyquist/Heyser spiral is rather akin to looking at a coiled slinky that is sitting in a stacked position from the end where you are looking into a coiled cylinder. Any point on the curve is represented by a vector, with the length corresponding to the amplitude and the angle from the real(x) axis corresponding to the phase angle. This corresponds to the complex impedance you may be familiar with calculating that consisted of a real number and an imaginary component with values in the 1st and 2nd quadrants being inductive, and the values in the 4th and 3rd quadrants being capacitive. So, for any point on the curve you have a unique real and imaginary value. But here is where this curve becomes truly amazing, and you begin to realize that you have MUCH more information then you imagined and hence why it is such an amazing source of information. If you take the end of the spiral and pulled it out of the paper along the z axis corresponding to time that extends 90 degrees perpendicular to the graph, you have a spiral curve that shows this response in the time domain. It is much like pulling the end of the slinky and extending the spiral. So far you have been looking straight down into the paper and only seeing the x and y axes and the real and imaginary components. But as we extend the spiral in time, we are not limited to looking at this spiral in 2 space! And here is where we begin mapping topologies between domains and the fun and insight really begins! Imagine that we have pulled this spiral out along the z axis perpendicular to the page. If you were to look down from above and project the spiral down onto a piece of paper, you have a representation of the impulse response! And if we looked at the same curve from the left side and projected it onto another piece of paper, you have the doublet response! But if you recall that we can map topologies between domains by such transforms as the Fourier transform we have MUCH more still! This same spiral viewed in the time domain provides the impulse response when viewed from above, and the doublet response when viewed from the side. But with the Fourier transform we can map this time domain topology to the frequency domain and the time domain impulse response maps to the frequency domain coincident response, and the time domain doublet response maps to the frequency domain quadrature response. Confused yet? The neat thing is to see this spiral in 3space, where you can move a curser along the curve and have the values at any point projected to the projection below and to the side! And this is exactly what the TEF20 allows you to do! (after it allows you to generate the response to begin with!) So it not only allows you to generate and quantify the response, it allows you to resolve it into its corresponding parts both visually and quantitatively! But it goes still further! If you recall that we are looking at the system response in both the time and frequency domains, the time domain phase response (the angle from the x axis) correlates to the instant frequency at that point, and the frequency domain phase correlates to the group delay! I guess you are probably overwhelmed by correlations and transpositions by now! When my TEF is shipped back to me (as a friend is currently borrowing it to shoot and align a new system in a studio in Nashville) I will try to capture a few screen shots of this and post them. But I hope that you can see that the nominal impedance becomes a very marginal value for a system that is dynamic with respect to the time and frequency domains. And this response can be determined for electronic devices as well as transducers/speakers. And we havent even begun to really use the TEF for the acoustic measurements! While MLSSA, SysID, SMAART and other systems allow you to measure various aspects (albeit without the noise immunity offered by Heysers time domain spectrometry system complete with swept frequency tracking filters) the TEF is truly the holy grail and the single most complete of all the analytical tools for acoustics. And it acts as a front end and compliments acoustical modeling programs like EASE, etc. But to recap, if one wants to optimize a source load pair couple according to the maximum power transfer theorem and thus minimize impedance mismatches resulting in reflections and resonance, a close correlation between source and load is necessary. And simply matching the nominal impedance at a particular frequency does not even come close to correlating a complex system. That is kind of like focusing on one particular play in a football game and pretending to be able to extrapolate the rest of the game from that one play. While convenient, would you be satisfied to see just one play of the Super Bowl, or just one pitch of the last game of the World Series and feeling content to claim an understanding of the entire game? Do you want to place a large bet based upon that simplistic bit of information? And regarding amps, there are other bits of very useful information such as the damping factor (the relative measure of its ability to decelerate and accelerate a dynamic driver in an EM field- especially for subs!), distortion figures, slew rates, etc., but given a well designed amplifier with reasonable specs, matching the source to the load is a more meaningful endeavor. After all, an amp that is capacitive in the lower several octaves is not optimal to drive a subwoofer. Remember that capacitors tend to dislike low frequency signal variations! And an amplifier that is inductive in the mid and or high frequencies is not generally an optimal match for driving mid and high frequency reproduction! On the other hand, an amp that is capacitive might be ideal for driving mid and high frequency loads, and an inductive amp might be ideal to drive low frequency loads. In other words, you have a more complete impedance signature from which to work. Well, I have rambled far too much. But I hope that this might perhaps stimulate a few thoughts about other more useful methods of evaluating components and systems! And we havent yet gotten to the real gold mine, that of analyzing the system within the environmental context of the acoustical time domain of speaker and room. Quote Link to comment Share on other sites More sharing options...
djk Posted January 16, 2005 Share Posted January 16, 2005 You really don't have a clue as to how speakers and amplifiers work. If you put an amplifier that matched its output impedance on a Cornwall it would have a 12dB peak in the midrange. Speakers are designed to be driven from a zero ohm source impedance. Quote Link to comment Share on other sites More sharing options...
Colin Posted January 16, 2005 Author Share Posted January 16, 2005 what does that mean? matched its output and create a mid-range bump? Quote Link to comment Share on other sites More sharing options...
Guest Anonymous Posted January 17, 2005 Share Posted January 17, 2005 i would really like to see a picture of this test, it sounds very interesting, but even so, you still have to use speakers to measure the output of this right? and from then wouldn't amny of these measurements be different with different speakers, just as if you had the same speakers and different amps? i understand completly what you are saying and your explination is clear, but correct me if i am wrong, but it seems that there are to many variables that would alter the spiral and the read outs, i mean changing a wire, change the amp, change the source, any of these? Quote Link to comment Share on other sites More sharing options...
MrMcGoo Posted January 17, 2005 Share Posted January 17, 2005 djk, I think that you have called it correctly. Damping factor has been known for many decades to provide very little useful information about damping. Any number over about 20 is good for damping. PWK has written an article that damping factor is really a good surrogate for output impedance. The higher the damping factor, the lower the output impedance. Low output impedance helps an amp achieve better frequency response as a speaker's impedance varies with frequency. Obviously there is more to frequency response than just looking at the output impedance. Bill Quote Link to comment Share on other sites More sharing options...
Guest Anonymous Posted January 17, 2005 Share Posted January 17, 2005 how come a speaker is rated at 4 or 8 or whatever, but it will vary why not say that this speakers range of frequencies will be from 1 to 10 ohms or something? is 8 or 4 the average? Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted January 17, 2005 Share Posted January 17, 2005 I will try to respond to the many comments some more worthy of comment then others! ;-) You really don't have a clue as to how speakers and amplifiers work. Well, thanks for your enlightened welcome to the Klipsch forum where so many decry the monopoly on arrogance and ignorance of marketing claims of intellectual black holes such as Bose and Monster! I would only debate the notion of a monolpoly! ;-) Yes, I guess I would agree, if by this you mean I dont have a clue relative to the understanding of such luminaries and mentors as Dick Heyser, Don Keele, Don & (the ever wonderful) Carolyn Davis, Dr. Gene Patronis, Bruce Howse, Russ Berger, and SO many others whos level of insight provide a standard by which to aspire and to be compared! I would submit that the primary difference between myself and so many with whom I have thus far encountered here is that despite what I do know (and no offense intended to those who feel that the study of physics simply complicates things!), I DONT think I know it all! I am simply a messenger. In fact, I am acutely aware of many things that I dont know! But then I dont simply dismiss concepts that dont fit into my limited knowledge either! I seek answers to these areas of ignorance! And regarding the Nyquist plot/Heyser spiral and its incredible wealth of contained knowledge, may I suggest that you debate the venerable Don Keele and others such as Don & Carolyn Davis and Dr. Gene Patronis. And I would suggest that you begin your quest with references such as Syn-Aud-Con, and Time Delay Spectrometry An anthology of the works of Richard C. Heyser on measurement, analysis, and perception AES publishers, ANY Don Keele speaker review, and The Theory and Design of Loudspeaker Enclosures by J.E. Benson. But I will defer to any of these sources if my description is not as complete as they could present! My inability to present such a complex system as clearly as I would like in such a limited space by no means indicates a deficiency in the method described! But I must take exception with those rocket scientists, who instead of asking questions, would dismiss that which they have NO clue about! It is rather amazing to hear that Dick Heyser, et. al., have no clue regarding the Nyquist plot! Sorry, I didnt catch your name! It wouldnt happen to be Lip****z or Vanderkooy would it? I assume that you think the pressure zone microphone to be folly as well?? Speakers are designed to be driven from a zero ohm source impedance. I am not quite sure where to begin or exactly how to respond without some saying I am being arrogant or condescending! While my wit is very dry and my comments generally said with a wry grin, and intended to be interpreted as such!!, may I simply say that I would genuinely like to see this animal to which you refer! Just one would suffice! ;-) But it is rather scary to think that many seem to think that simple DC resistance provides the entire story for a dynamic system operating in the AC realm! Funny, when I look in the index of many books I am faced with MANY definitions of types of impedance! You mean there isnt JUST one? Hmmm. If you put an amplifier that matched its output impedance on a Cornwall it would have a 12dB peak in the midrange. A 12 db peak? I dont know what profound aspect of the theory to which you respond. I believe that you are misunderstanding my intent in speaking about matching a load and a source. I doubt any matching of source and load per se is going to correct for the total loads response. The purpose of this information is not to solve all such anomalies! But the optimal matching will further expose the response characteristics of the load, however profound or deficient. I am not sure where you are going here you still have to use speakers to measure the output of this right? To measure the response of the load you can measure the response of individual components or of the summed components. Personally, while an individual components value can be very useful, the most useful measurement of a load is the total load including interconnects, crossovers, transducers, etc the TOTAL load (aside from some of the attitudes ;-). After all, arent so many making some extraordinary claims regarding the impact of cables, cosmic radiation, sunspots, barometric pressure, and your favorite color of M&M (is that peanut or regular? Hmmm, so many variables!) upon the total load and hence the response of a system? ;-) The Nyquist plot is but one VERY powerful tool for matching a source with a load! It does not obviate comprehensive design of each component according to best practices employing proper diligence! I would and from then wouldn't any of these measurements be different with different speakers, just as if you had the same speakers and different amps? Would different speakers present different loads? And would different amps present different source impedance signature (etc!) over the entire frequency range? Sure! Sure, as would different components which comprise the total load change the toal load signature! But before anyone jumps on this to prove some other point, certain components (such as transducers!) tend to have the largest impact upon the response! There is NOT ONE absolute Nyquist response! They are unique for each source and load. Hence the value in measuring and determining them! i understand completely what you are saying and your explanation is clear, but correct me if i am wrong, but it seems that there are to many variables that would alter the spiral and the read outs, i mean changing a wire, change the amp, change the source, any of these? I think I mentioned that above. Of course, but each component has a relative effect on the total response! At the risk of being a smart aleck (who me?!), the response of the interconnect generally does not take precedence over the transducer, despite the fact that some will spend almost as much on the cable as they do the speaker! ;-) Any number over about 20 is good for damping. PWK has written an article that damping factor is really a good surrogate for output impedance. The higher the damping factor, the lower the output impedance. Low output impedance helps an amp achieve better frequency response as a speaker's impedance varies with frequency. A quick response to this iswell, many concepts seem to be combined and oversimplified here Damping is (in our case) a function of the electromotive force enabling the change in a magnetic field to mechanically damp control the movement of a voice coil within the magnetic field - to effectively stop and accelerate said element. While generalizations can be made, the effects of damping for an electrostatic element are significantly different from a bass reflex subwoofer where the driver is mechanically un-damped at its free air resonance (and brick wall filtering not employed!!) And I know for a fact that Paul held Dick Heysers work in the highest regard, so I dont think you will be able to use Paul to refute Heysers concept! And as regards the frequency domain, may I suggest that you need to understand the role of the precedence of the time domain response from which the frequency domain is but another frame of reference. I will defer to Dick Heyser, Don Davis and Dr. Patronis for this one! But, I believe that you are mistaking the relationship between impulse response and the frequency response, where the smaller(shorter time) of an impulse response maps to a wider frequency response, and where, at least theoretically, a zero time impulse response results in an infinite frequency response. Although, I have never heard this! Maybe I need to get some of the magic cables that some swear by! Obviously there is more to frequency response than just looking at the output impedance. Obviously? May I suggest that you study the relationship between frames of reference and the mapping of topologies from one domain to another domain. If you are treating them as distinctly different and unique realms, you have a fundamental misunderstanding of the integral relationships to which Dick Heyser spoke. how come a speaker is rated at 4 or 8 or whatever, but it will vary why not say that this speakers range of frequencies will be from 1 to 10 ohms or something? is 8 or 4 the average? Ah, perhaps because that is what a particular herd of cats belonging to the IEEE decided at a standards conference? Why do we say a baseball player is hitting with an average of say, .296? Does that mean that every time he bats that he will perform with this same probability? Can he have a day when he hits .000 or 1.000? It is a generalization where MANY variables are ignored for the purposes of convenience! Most notably, the time domain is ignored and the value is treated as a static variable for convenience! But all generalizations are by definition oversimplifications! Does that mean they have no value? Of course not! They are used precisely for their convenience! But it is a mistake on the part of the individual to ignore the implied underlying complexity! A number of references: http://www.svconline.com/mag/avinstall_response_different_types/ And be sure to look at this!!!! Not only are their diagrams of the output, but many more articles that may help http://www.prosoundweb.com/install/sac/n26_4/nyquist/nyquist.php and dont miss Pat Browns article in the same set of references on the amplifier-speaker interface! Note the components of the load and why the complex impedance provides much more info than the simple time-invariant resistive component! Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted January 17, 2005 Share Posted January 17, 2005 And why is the name of "L i p s h i t z" asterisked out? Granted, I don't have much respect for his views, but it IS his name! Its a sad comment when even Klipsch treats this forum as if everyone is a 5 year old. What does that imply regarding its marketing attitude? Does that mean they also assume all are completely ignorant of all things electro-acoustic as well as some other companies do as they obsfucate issues with their marketing techno-babble? Hmmm? Amazing! Quote Link to comment Share on other sites More sharing options...
yaffstone Posted January 17, 2005 Share Posted January 17, 2005 Wow, this thread has some flame power! Not having gone to tubes yet, I have relied on numbers in the past when buying SS componenets. My theory has been to keep the source signal unaltered as much as possible by the intervening components like the pre-amp and amp. Therefore, I look for units which are rated at 0.05% THD or lower and high S/N ratio (>= 100 dB). I know there's a lot of argument about the THD of the first 1/10 of a watt and I don't pretend to answer these concerns. All I can say is that this approach has worked well for me. I have been very happy with my limited equipment purchased this way. The list includes: Crown PSL-2 stereo pre-amp, Kenwood M-1 amps (now dead and gone) and most recently Outlaw Audio amps and preamp. All of these units have sounded great with Klipsch speakers, negligable hiss at zero volume (one has to put their ear in the speaker to hear any) and great sound at listening volumes. In the end, you have to give a listen and decide what you prefer. Quote Link to comment Share on other sites More sharing options...
Dylanl Posted January 17, 2005 Share Posted January 17, 2005 Dragon, you sure know how to lengthen a thread. Since I have been on this forum I have never seen explainations that long. I started to fall asleep reading the 1st one. 2nd one ZZZZZZ! Huh, what was that I was reading?? 3rd Good night - ZZZZZZZZZZZZZZ Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted January 17, 2005 Share Posted January 17, 2005 Gee, I sort of felt the same way when they said I had to take all those physics classes when I was in school! But then if I hadn't taken the time to pay attention, I too might be out there paying up to $1500 for a meter interconnect and arguing about it here still as so many others are! ;-)) Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.