I second the comments that have been made by several people to the effect that the answer to the original question is speaker dependent. Winggo, I see that in a thread you started about a year and a half ago you mentioned having a pair of AR9 speakers. If those are the ones in question, given their vintage (1970s), their modest 87 db efficiency, their nominally 4 ohm impedance, and their high power handling capability, it is safe to assume that they were designed with solid state amplification in mind. As some of the others have said, the suitability of using high powered tube amplification will be significantly dependent on how the impedance magnitude and impedance phase angle of the speaker vary as a function of frequency. I couldn't find any indication of the speaker's phase angle characteristics, but according to this review their impedance magnitude is generally between 3 and 5 ohms, with peaks of 10 ohms at 750 Hz and 8 ohms at 27 Hz. In comparison with a solid state amp, a tube amp will tend to give slightly more emphasis to frequencies in those vicinities. Another point I would make is that IME a forte of tube amplification tends to be imaging and dimensionality. Although I have no basis upon which to be certain, my suspicion is that that is not a strong point of these speakers, and that potential benefit of tube amplification would likely be wasted on them. My bottom line guess is that these speakers would do reasonably well with something like one of the higher powered Rogue Audio tube amplifiers, but you would be likely to do just as well if not better with a somewhat lower priced solid state model. Regards, -- Al |
05-10-13: Mattmiller
Efficiency and Sensitivity are words that get confused when talking about speakers. That is true. But I don't think that the references to efficiency in several of the earlier posts represent unreasonable uses of the term. Strictly speaking, speaker efficiency would be defined as acoustic power out vs. electrical power in. It would be expressed as a percentage, and would be a very small number for nearly all speakers. It would also be a number that is rarely specified, and that is not particularly helpful. It is common, though (and reasonable, IMO), for the term "efficiency" to be used to refer to sound pressure level at a distance of 1 meter in response to an input of 1 watt. And for "sensitivity" to be used to refer to SPL at 1 meter in response to an input of 2.83 volts. 2.83 volts corresponds to 1 watt into 8 ohms. So for an 8 ohm speaker "efficiency," per that loose definition, and "sensitivity" would be numerically identical. For a 4 ohm speaker, efficiency per that definition would be 3 db less than sensitivity, since 2.83 volts into 4 ohms corresponds to 2 watts. That difference will have greater significance in the case of a tube amp than a solid state amp, because the power capability of a tube amp into 4 ohms will generally be similar to its capability into 8 ohms, while the power capability of a solid state amp into 4 ohms may be as much as twice its capability into 8 ohms. On another note, in my previous post when I said: Another point I would make is that IME a forte of tube amplification tends to be imaging and dimensionality. Although I have no basis upon which to be certain, my suspicion is that that is not a strong point of these speakers, and that potential benefit of tube amplification would likely be wasted on them. it occurs to me that my wording could be taken more negatively than it was intended, due to my use of the word "wasted." To clarify, my intention was just to say that I suspect that the potential benefit that IMO tube amplification can often provide with respect to imaging and dimensionality would probably not be realized with AR9 speakers. One reason for that suspicion, in addition to its general configuration, is the comment in the review I linked to that the AR9 is designed to be placed close to the wall, which usually results in imaging being compromised to some degree. Regards, -- Al |
Hi Bruce,
I suspect that the specified damping factor of 40 for the MC-352 (actually, it is specified as "greater than 40," but I'll assume 40 for present purposes) represents the output impedance of the amp for a given output tap divided into the speaker impedance that tap is nominally intended to be used with.
If so:
The output impedance of the 8 ohm tap would be 8/40 = 0.2 ohms.
The output impedance of the 4 ohm tap would be 4/40 = 0.1 ohms.
The output impedance of the 2 ohm tap would be 2/40 = 0.05 ohms.
The output transistors would presumably see the same load when an 8 ohm speaker (that is truly 8 ohms) is connected to the 8 ohm tap, compared to when a 4 ohm speaker (that is truly 4 ohms) is connected to the 4 ohm tap, compared to when a 2 ohm speaker (that is truly 2 ohms) is connected to the 2 ohm tap.
BTW, thanks for your kind comment earlier in the thread.
Best,
-- Al |
Thanks very much, Matt. Yes, the Daedalus Ulysses, in the configuration I have that was introduced about 3 years ago which includes "all poly" crossover capacitors, is 98 db/1W/1m. It also has a very flat impedance curve, as well as high power handling capability, all of which makes it unusually versatile with respect to amplifier selection.
Regarding your question, what I said was "the power capability of a solid state amp into 4 ohms MAY [emphasis added] be as much as twice its capability into 8 ohms."
As you probably realize, the power capability of MOST solid state amps will be substantially greater into 4 ohms than into 8 ohms, in many cases approaching twice as much, and in some cases equaling twice as much. Two ohms may be a different story, of course, as a lot of amplifiers will not do well dealing with such a heavy load.
As you indicated, though, your McIntosh MC-352 amplifier has the same output rating into 8, 4 and 2 ohm loads. That is essentially a consequence of the unusual autoformer-based design of its output circuits. I am not aware of any non-McIntosh amplifiers that use a similar approach.
As opposed to a tube amplifier utilizing an output transformer, however, the solid state + autoformer approach provides considerably lower output impedance/higher damping factor (damping factor being spec'd at 40). As a consequence of that it WILL, like most solid state amplifiers, to a very close approximation double actual power DELIVERY (as opposed to CAPABILITY) as load impedance is halved, AS LONG AS it is operated within its 350W or so capability, and as long as the impedance of the speaker is a reasonable match for the output tap on the amplifier that the speaker is connected to.
In contrast to most tube amps, that essentially places the MC-352 in what Atmasphere refers to as the voltage paradigm camp, i.e., its output voltage for a given input will be essentially constant regardless of load impedance, as long as the amplifier is operated within its capabilities.
Regards,
-- Al |
Excellent comments by all, IMO. Charles, re the volume levels I listen at, Wolf's comment that: 05-21-13: Wolf_garcia
Hearing issues come from sustained high levels, and unlike something like a gunshot next to your head, a short musical burst (from classical music for example) isn't going to hurt you usually is exactly applicable. My average listening level on most classical music is probably in the mid-70's. But I have, for example, a great many of the Telarc digitally mastered vinyl releases from the 1980's, which employed zero compression of dynamic range. Many of their symphonic recordings were notorious for their very loud bass drum beats, which can easily exceed 100 db at my listening position with the volume control set such that average levels are completely safe. Also, as I mentioned in a previous thread here: One of the widest dynamic range recordings in my collection (although there are many others that are close) is the Sheffield Labs recording of Prokofiev's "Romeo and Juliet," conducted by Erich Leinsdorf. Out of curiosity, a while back I examined its waveforms on a computer, using an audio editing program. The difference in volume between the loudest notes and the softest notes was around 55 db!
On that recording, and quite a few other classical symphonic pieces I have on high quality labels that tend to use minimal or no compression, I frequently measure brief peaks at my listening position in the 100 to 105 db area, using a Radio Shack digital SPL meter, although the average level is by no means particularly loud.
Those peaks are not nearly as loud, btw, as those I heard when I once listened from the very front row at Tanglewood to the Boston Symphony Orchestra performing that same Prokofiev work. My guess is that the peaks easily reached 115 db. Regarding the definition of "speed," many good comments have been made. If I were to summarize my understanding of the term in just a few words, I would say it means that fast musical transients are reproduced ACCURATELY. 05-21-13: Csontos
Judging from your responses it's fairly clear that it is in fact a technically quantified attribute and therefore realizing it is not subjective at all. The reason I don't care much about specs is because amps don't necessarily live up to them. But I am aware of which ones they are. I wouldn't go as far as to say that "speed" is technically quantifiable. The bandwidth specs I referred to earlier of course tell just a small part of the story. Bandwidth extending several-fold beyond the limits of our hearing is necessary to maintain phase relationships between harmonics that are at frequencies in the upper treble region, and lower frequencies that will be present in the same musical notes. But that and other speed-related technical parameters, such as slew rate, certainly don't tell the whole story IMO. For instance, it is well established that excessive use of feedback is conducive to Transient Intermodulation Distortion (TIM), which as its name implies will, if present to a significant degree, result in inaccurate and sloppy reproduction of fast transients. As far as I am aware, TIM is never specified, and I'm not sure that standards defining how it should be measured even exist. How to minimize or eliminate TIM? Design an amplifier that doesn't need much or any feedback, such as many tube amps and some (but relatively few) solid state amps. Best regards, -- Al |
05-20-13: Csontos
I would venture to guess that Al, but especially Ralph believe it's not possible to arrive at life-like sound without tubes given some of their prior posts. To that end I would like very much for either/both to cite some tube amps that are as fast as any ss amps they've heard so far. Imo, you can't get there without serious speed in an amp. Hi Peter, No, I for one would certainly not make any such claim. In part because my listening experience with solid state amps during the past decade or two has been minimal. Prior experience led me in the direction of tubes, with which I've been very happy, and in recent times I've simply had no interest in high end solid state. And my speaker preferences have always tended to be in the direction of those having impedance characteristics that are benign and tube-friendly. For examples of tube amps that are "fast," for starters consider Ralph's OTL amps. Although I haven't had the pleasure of hearing any of them, by all accounts I've ever seen they have that characteristic in spades. And among tube amps having transformer coupled outputs, my VAC Renaissance 70/70 Mk III certainly doesn't disappoint. While I realize that you "don't care about published specs," I'll mention that they are rated as being down only 0.5 db at 90 kHz on a small signal basis, and at 58 kHz at full power. At -3 db those ratings are 103 kHz and 85 kHz, respectively. Not very different than the very highly regarded Pass XA.5 series solid state amps, which are rated to 100 kHz (with no specification of power level or how many db down that is based on). Yes, I know, a few solid state amps, such as those made by Spectral, go out to 1 or 2 MHz or more, but whether or not that is overkill, and whether or not it may even be sonically counter-productive, are certainly debatable. As a point of reference, I'll add that I frequently compare listening results between my amp + speakers and my Stax electrostatic headphones + Stax solid state headphone amp. I doubt that anyone would ever accuse the Stax components of being "slow," yet I am not disappointed with the "speed" of the amp + speaker combination. Tubegroover, thank you most kindly for your expression of appreciation. Best regards, -- Al |
Hi Charles,
Yes, the VAC Renaissance 70/70 Mk III has a six-position switch for each channel which, according to the manual, allows the amount of feedback to be selected between 0, 3.5, 4.5, 5.5, 6.5, and 7.5 db (the markings on the panel omit the fractional amounts).
I've been sufficiently happy with the results provided by the 0 db setting, though, that I've never even tried the other positions. If the impedance curve of my speakers were not as flat as it is, however, and if their impedance characteristics were less benign in any other way, I certainly would have felt motivated to experiment with the other settings.
Best regards,
-- Al |
Hi Matt, I have no experience with Dynaudio speakers, but my instinct would be to proceed with caution before going to tube amplification with them. In addition to the comments by Ralph and Mapman just above, I would emphasize that high quality high powered tube amplification tends to be quite expensive. And you clearly need lots of watts. Based on the impedance curve of your speakers, shown in Stereophile's review here, I would consider their 86 db/2.83V/1m sensitivity to be roughly equivalent to 84 db/1Watt/1m. At typical listening distances (say 10 to 12 feet or so) that would mean that 70 watts into both speakers would produce a maximum SPL of around 96 db or so. While that is adequate for most listeners when listening to highly compressed rock music, I suspect it would be very disappointing with a lot of material having wide dynamic range (i.e., a wide difference in volume between the loudest and the softest notes). My listening is about 90% classical (including a lot of wide dynamic range symphonic material), 5% rock, and 5% miscellaneous. I've measured some of the classical material I listen to as reaching 105 db at my listening position, which my 98 db/1W/1m speakers and 70W amp handle comfortably. Although I haven't heard them, if you do want to consider tube amplification you may want to look into the higher powered models from Rogue Audio. Some of them provide close to 200 watts, at relatively reasonable prices, and seem to receive generally favorable user comments. Best regards, -- Al |
05-21-13: Bifwynne
Al, kinda curious about the variable negative feedback switch you mentioned. Out of curiosity, if you were driving a speaker with a bumpy impedance curve, how much "smoothing" would be achieved if you used NF?? For example, if your speaker had a 20 ohm bump at 2K Hz, what would happen if you used the maximum amount of NF? Hi Bruce, I don't know. It would depend on the particular amplifier, and what its output impedance is with and without that amount of feedback. Also, keep in mind that while what would be smoothed as feedback is increased is the variation of output voltage as a function of variations in speaker impedance over the frequency range, the variation of how much power is delivered to the speaker as a function of speaker impedance variation would probably be worsened. And the net sonic effects of all of that would be dependent on the design of the particular speaker. Another question out of curiosity. What is considered a large amount of NF? For example, my amp puts out a pretty typical 26db or so of gain, and uses about 12 db of NF. What does that mean in terms of magnitude, output smoothing of speakers with bumpy impedance loads and the introduction of TIM? Ralph would most likely have a better quantitative feel for that than I do. But my general impression is that 12 db of feedback is probably a bit more than average for a high-end tube amp. Re TIM, variables other than the amount of feedback, such as propagation delay through the particular amplifier (i.e., how much time is required for an input signal to be propagated to the output, which is a parameter that is rarely if ever specified) will have major effects on the answer. Best, -- Al |
05-24-13: Mapman
... I believe Lincoln Walsh used his engineering experience with radar systems in WW2 to get dynamic speakers right while the rest of the world missed the boat. Pretty radical. Interestingly, as far as I am aware he was also the only manufacturer of consumer-oriented audio power amplifiers utilizing the 300B directly heated triode power tube, between the earliest days of hifi and the resurgence of interest in that tube that occurred in the last three decades or so. Also, some of his amplifiers used 2A3 DHT power tubes, and similar comments apply to them, with the slight qualification that his use of 2A3s was preceded by their use in the power amplifier sections of certain high end radios of the 1930s. Those Brook Electronics Company amplifiers are highly collectable these days, and go for big $$$ on the rare occasions when they appear for sale. Best, -- Al |
