I have no particular question to ask at this time, Roger, but I want to take the opportunity to express my appreciation for your participation here. As with Ralph (Atmasphere) and a few other audio designers who participate here at least occasionally, I always follow your posts with interest, and I have learned from them. And I’ve learned from them even though I have BSEE and MSEE degrees myself, from excellent schools, as well as 33 years experience designing and managing design of sophisticated analog and digital circuits (not for audio), and nearly 40 years of experience as an audiophile.
Thank you, and best regards,
-- Al
|
There is a lot to this question so lets hear back from you and others.
My own philosophy is based on the expectation, generally speaking, that for a given level of quality, and within a given architecture (solid state, tube push-pull, tube SET, etc.) and class of operation (A, AB, D, etc.), there will tend to be at least a loose correlation between amplifier power capability and amplifier cost. And in choosing an amplifier I prefer to direct as high a percentage as possible of the amount I want to spend toward quality, rather than toward watts. My listening includes a lot of well engineered classical symphonic recordings, which therefore have very wide dynamic range, and not infrequently reach brief dynamic peaks approaching 105 db at the listening position. So I choose an amplifier that in conjunction with my speakers can comfortably support that level, with a few db to spare, and I don’t pay for any more watts than are necessary to accomplish that. Instead, I try to direct whatever $ I choose to invest in an amplifier toward quality, as much as possible and as best as I can determine that based on research and listening. On another note, happy Thanksgiving to all! Best regards, -- Al |
Ramtubes 11-22-2018
@almarg
It would help to quote the original question so we can follow this.
It appears to be an answer but im not getting it. Roger, my previous post was a response to your post that immediately preceded it, about the need or lack thereof for large amounts of amplifier headroom. And implicit in my post was agreement with your earlier statement that "over 100 watts is only justified by either high listening levels or insensitive speakers or both together. Excess headroom is a myth," at least as far as I am concerned. Best regards, -- Al |
Ramtubes 11-26-2018:
HI, Roger here with a question.
I would like to hear how each of you figured out how much power you needed to buy in watts?
I would appreciate the following information in your response.
Your listening level LOUD SPL (preferably measured at 1 meter from the speaker with a REAL SPL meter. Your low listening level. If you are using a cell phone app then you have confirmed it?
Your speaker sensitivity?
Listening Distance from speakers?
How many watts at your load is the amp is rated to supply?
As I mentioned earlier in the thread I listen to a lot of classical symphonic music that has been engineered with minimal or no dynamic compression. Two such recordings, which I believe have just about the widest dynamic range in my collection, are Stravinsky’s “Firebird Suite” on Telarc (Robert Shaw conducting the Atlanta Symphony), and Prokofiev’s “Romeo and Juliet” (excerpts) on Sheffield Lab (Erich Leinsdorf conducting the Los Angeles Philharmonic). I have examined the waveforms of those recordings using a professional audio editing program (Sound Forge Pro), and by doing so I have found the difference in volume between their loudest notes and their softest notes to be approximately 55 db, which is (to me) amazing. Correspondingly, at my 12 foot listening distance I have measured peak SPLs on those recordings of close to 105 db, with the softest notes being in the vicinity of 50 db. I used a Radio Shack digital SPL meter for these measurements, set for C-weighting and fast response. My speakers (Daedalus Ulysses) are rated at 97.5 db/1w/1m, and have a very flat impedance curve with a specified nominal impedance of 6 ohms. My 12 foot listening distance corresponds to 3.66 meters. Putting aside room effects for the moment I assume that SPL produced by a box-type dynamic speaker such as those falls off at 6 db per doubling of distance, which means an 11 db reduction going from 1 meter to 12 feet. I conservatively add in 3 db to reflect the presence of two speakers (as I understand it that figure will actually be closer to 6 db at my centered listening position when both speakers are producing similar signals), and I add in perhaps 3 db for “room gain.” 97.5 -11 + 3 + 3 = 92.5 db at the listening position for 1 watt per channel. Let’s call it 93 db. I add in about 3 db of margin to the 105 db I want my amp/speaker combination to be able to produce at the listening position. So the required minimum amplifier power (into 6 ohms) is: 105 + 3 - 93 = 15 dbW (decibels above 1 watt) 15 dbW = 32 watts. To answer your question about my amplifier, for several years prior to just recently I was using a VAC Renaissance 70/70 MkIII, rated at 70 wpc. I recently changed to a Pass XA25, which is specified as a class A amplifier rated at 25 wpc into 8 ohms and 50 wpc into 4 ohms. Per JA’s measurements, though, it is capable of 80 and 130 wpc into those impedances. I presume that most of that increase represents the amp’s capability after leaving class A, although per JA’s comments some of the increase apparently reflects differences in the distortion percentages the ratings and measurements are based on. Best regards, -- Al
|
Ramtubes 11-27-2018:
Thanks for your measurements and math. I think the math will leave most with their head swimming. Perhaps you could add some details to the steps so that others, less math inclined, might work things out. it is nice for people to know how different speakers fall off with distance. Perhaps you could write that up for us. :) I ask for SPL both at listening position and 1 meter for two different purposes. I want to get to know the listener and measuring at 1 meter leaves out all those other calculations. One just works off the 1 meter speaker spec and adds or subtracts. If one is 3 db higher than the speaker spec he is at 2 watts, 6 db is 4 watts, 10 db is 10 watts, 20 db is 100 watts. Thanks for your comments about my post, Roger. I guess the first thing I should emphasize is that the dynamic ranges of the two recordings I referred to in my previous post, while being applicable to MY determination of minimum required amplifier power, are VASTLY greater than the dynamic ranges of most or all of the recordings most others listen to. And in fact are much greater than the dynamic ranges of most of the recordings I listen to. So my bottom line of a 32 watt minimum power capability, for my purposes, is likely to be considerably less for those who listen at comparable average volume levels (average SPLs say in the 70s at the listening position) **if comparably efficient speakers are used.**
This is consistent, btw, with statements I believe Roger made earlier in the thread to the effect that many listeners require less power than they tend to believe. I’ll also say that the point to going through the kind of calculations I described in my previous post is of course NOT to make any kind of final determination of what amp to choose, which of course should be made by listening, assuming that is possible. The point is to narrow the field of candidates that are to be considered, and to minimize the likelihood of making an expensive mistake, that would perhaps work for some of the listener’s recordings (especially those having narrow dynamic range), but would not work (or at least would not work well) for other recordings (especially those having wide dynamic range).
Another important point to keep in mind is that speaker sensitivities are often specified as the SPL produced at 1 meter in response to an input of 2.83 volts, rather than 1 watt. For an 8 ohm speaker (that is truly 8 ohms) 2.83 volts corresponds to 1 watt, so the two kinds of specifications will be identical. But 2.83 volts into 4 ohms corresponds to 2 watts, so for example a 90 db/2.83 volt/1 meter/4 ohm speaker is really just an 87 db/1 watt/1 meter/4 ohm speaker.
Regarding measurements at 1 meter, I haven’t done that (aside from during the unique speaker calibration processes that are required to utilize the capabilities of my DEQX HDP-5), mainly because it wasn’t necessary for my purposes, and because at least in my case making the determination from the listening position and performing the necessary calculations isn’t difficult. Also, while 1 meter measurements would simplify the calculations somewhat, as Roger mentioned, one would still have to somehow address the effects on SPL of having two speakers, as well as addressing room effects in some manner. ... it is nice for people to know how different speakers fall off with distance. Perhaps you could write that up for us. :) In general the SPL provided by a box-type dynamic speaker (meaning a speaker that is non-planar and is not a line source) will fall off at 6 db per doubling of distance, **neglecting the effects of room reflections.** (As can be seen in my previous post, I added 3 db to my calculations as a rough guess of "room gain" as perceived at the listening position). The fall-off of planar and line source speakers as distance increases will be significantly less than that, as those speakers tend to throw their sound forward. The reason being, I believe, that as listening distance increases the angle between the listener’s ears and central part of the speaker and the upper part and the lower part of planar and line source speakers decreases to a greater degree (pun intended) than in the case of typical dynamic speakers having drivers that are spaced relatively closely. Making the sound emitted by those parts of planar or line source speakers more able to contribute to the perceived SPL than at smaller distances. A fall-off of 6 db per doubling of distance can be extrapolated to any combination of distances based on the formula 20 x log(D1/D2), where “log” is the base 10 logarithm. If D1 is greater than D2 the answer will be a positive number; if D1 is less than D2 the answer will be the same number except with a minus sign in front of it. One can perform this calculation with a scientific calculator, or with the calculator that is built into Windows if it is set to scientific mode, or with various online calculators. In my case my listening distance is 12 feet. 12 feet is 144 inches; 1 meter is 39.37 inches; 144/39.37 = 3.66 meters. 20 x log(3.66/1) = 11 db, rounded off. So the SPL produced by a single box-type dynamic speaker at 12 feet will be about 11 db less than at 1 meter, neglecting room effects. The conversion of dbW to watts that I showed in my previous post is based on the ratio of two power levels, expressed in db, being 10 x log(P1/P2), where “log” is again the base 10 logarithm. If P1 is greater than P2 the answer will be a positive number; if P1 is less than P2 the answer will be the same number except with a minus sign in front of it. So it can be calculated that 32 watts is 10 x log(32/1) = 15 db greater than 1 watt, which can be expressed as 15 dbW. Converting in the opposite direction (15 dbW to 32 watts) is a bit trickier mathematically, but no doubt there are online calculators which can facilitate that. Best regards, -- Al |
Hi Ian (Ieales),
Surely you realize that pi squared, rounded off slightly, equals:
3.14 x 3.14 = 9.86
Best regards,
-- Al
|
Ralph (Atmasphere), thanks for your characteristically very constructive and informative post above. A very minor correction, though, for the record. I’m sure you misspoke when you referred to pi-squared. The 6.28 factor you referred to is of course correct, but that is 2 x pi, not pi-squared.
Thanks again. Best regards,
-- Al
|
@prof, I'm sure Roger would take this into account, but keep in mind that since the CJ Premier 12 and the Eico HF81 are both tube amps and have output transformers, they should not be operated unloaded while they are being provided with a signal. So a suitable switchbox would apply load resistors to whichever amp is not selected.
Best regards,
-- Al
|
I recall that in a thread here a while back Kevin Deal of Upscale Audio also mentioned very emphatically that 6SN7s are often microphonic.
FWIW, though, during the approximately seven years in which I owned a VAC Renaissance 70/70 MkIII, which uses four of those tubes, at various times I used a total of about 20 of them. All of them were the GTB version, with the majority being NOS tubes of various makes from the 1950s and 1960s, and the rest being current production. None were microphonic initially, as determined by lightly tapping on them with a pencil eraser. Two of them eventually became severely microphonic, however.
Best regards,
-- Al
|
Regarding the issue of cartridge loading, in a past thread here Jonathan Carr (Lyra cartridge designer, who I believe has also designed some phono stages) stated as follows: I should now debunk another myth regarding loading, which is that low-impedance MC cartridges are insensitive to capacitive loading. OK, the MC cartridges themselves aren’t particularly sensitive to capacitance, but the inductance of the cartridge coils will resonate with the distributed capacitance of the coils and the capacitance of the tonearm cable to create a high-frequency spike, and this spike certainly is sensitive to capacitance. In general, the less the capacitance the better. Having more capacitance (across the plus and minus cartridge outputs) will increase the magnitude of the high-frequency spike and lower its frequency, neither of which is good news for phono stage stability or phase response.
Generally speaking, the greater the capacitance across the plus and minus cartridge outputs, the heavier the resistive loading needs to be to control the resulting high-frequency spike. Conversely, less capacitance allows the resistive load on the cartridge to be reduced, which will benefit dynamic range, resolution and transient impact. The relevance of the Hagerman link Ralph provided in his previous post is that it illustrates the high frequency resonant peak (i.e., the "spike") formed by the interaction of the inductance of the cartridge and the load capacitance that it sees. I am not in a position to say whether or not the **only** relevance of resistive loading of a low output moving coil cartridge is to control that peak, and its potential effects on the particular phono stage. However I certainly wouldn’t consider the possibility that it could often at the very least be an important factor in the performance of a system to be "out of the world," as Roger stated. Also, while I don’t recall the exact numbers, Ralph has stated in some past threads that he has observed remarkably high levels of energy emanating from LOMC cartridges at ultrasonic or RF frequencies. Also, FWIW, I’ll mention that Keith Herron, whose company and products (especially his phono stage) are about as non-controversial and highly regarded as they come, suggests that with his particular phono stage no loading whatsoever will often be found to be preferable with LOMCs, regardless of the cartridge type. (The LOMC input of his phono stage is FET-based, and it applies a load resistance to the cartridge that is nearly infinite when load resistors are not connected externally, to RCA jacks that are provided for that purpose). And I have found that to be the case in my own system, with an AT-ART9 cartridge having a recommended load of "100 ohms minimum." Finally, without placing blame (although I have my own thoughts about that) I’ll just say that it’s a shame that an otherwise wonderful thread is compromised by the fact that two long-time designers of highly regarded audio electronics can’t deal with each other in a more respectful and matter-of-fact manner. Regards, -- Al |
Ramtubes 12-6-2018
...BTW Saul told me that around the shop they firmly preferred a pair of Mono 8s to the famous Model 9. That’s interesting to hear, Roger. Although to be precise I’m pretty certain that the 8 and 8B were both single-chassis stereo amps. I’ve never heard an 8 or 8B but during the 1990s I owned a pair of Model 9 monoblocks and a pair of Model 2 monoblocks. And I greatly preferred the sound of the 2s to the 9s, although in fairness I can’t exclude the possibility that condition may have been a factor in that. The 2 was similar in some respects to the 8 and 8B, which came later of course. Although in addition to being monoblocks the 2 employed tube rectification, while the 8 and 8B used solid state rectification, and the 2 used 12AX7 input tubes while the others used 6BH6s. I’ve commented here in the past that at least when used in conjunction with speakers having benign impedance characteristics and not requiring more than the 18 or 20 watts or so that the 2s were capable of in triode mode, in that mode the 2s were one of the best sounding amps I’ve ever heard. Regards, -- Al |
What would your phono stage be, hopefully I can find it’s output impedance .... George, I was wondering the same thing, so I took a look at Krelldreams’ posts earlier in the thread and it appears he is presently using a Schiit Mani phono stage (op-amp based, 75 ohm nominal output impedance, $129), but is looking into alternatives. While I understand that the Mani has received a lot of praise, and is certainly considered to be an excellent value at its very low price point, I can’t help but wonder if his finding that... The system with the passive has clear high and mid frequencies, good space, and sounds spacious... but it is a bit brighter, a bit leaner, and is less pleasant to listen to than with the preamp. I’d call the sound with the preamp “smoother” and “warmer”. The vocals through the preamp were slightly veiled compared to the passive though, which annoyed me. ... is simply due to the output of the passive preamp accurately reflecting what the Mani is providing to it. So it seems to me that it may be wise for Krelldreams to defer addressing the preamp question until an upgraded phono stage has been purchased and is in use. Best regards, -- Al |
gpgr4blu 12-12-2018:
I have enjoyed posts by Ralph and Ramtubes on this forum. They are both highly qualified as designers and manufacturers. They both provide answers to technical questions about gear. Sometimes, they include opinions based on their understanding of the science behind gear. Nevertheless, their posts are almost always informative to the membership here. Occasionally, they are self referential....
... I suspect the line should be drawn somewhere in between the services provided by Ralph and Ramtubes and the disservice provided by [redacted by Al, only because it is not relevant to this thread]. Might I add that Ralph and Ramtubes are not close to that line IMHO.
+1. Well said, gpgr4blu. And to the extent that Roger or Ralph might ultimately derive some pecuniary (monetary) benefit from their contributions here, IMO it would amount to a win-win. A win for the members here who benefit from the knowledge they share with us, as well as for them. Roger, I have a sincere question. What specific technical considerations lead you to be so negative about Ralph’s paradigm paper. Obviously suitability for use with a wide range of speakers is not a priority with his amplifier designs, as you’ve stated it is with your designs. But the only other relatively minor issue I’ve ever perceived in his paradigm paper is that as worded it might lead **some** readers to believe that the high output impedance and other characteristics of his amplifiers (and various other tube amplifiers) would result in precisely constant power delivery into varying load impedances. (In fact I’ve seen one or two posts by members here who do not have significant technical backgrounds in which that belief has been stated). Whereas the reality is simply that they will come considerably closer to accomplishing that than an amp which acts as a voltage source. To a greater or lesser degree depending on the amp’s output impedance and on how the speaker’s impedance varies as a function of frequency. I also think it’s noteworthy that some of Nelson Pass’ First Watt amps have even higher output impedances, and consequently "poorer" output voltage regulation, than Ralph’s. In some cases vastly greater output impedances than Ralph’s designs. But it seems to me that in both cases that doesn’t mean their amps are flawed in either concept or execution, it just means that they are intentionally designed such that they are suitable for use with a relatively small subset of available speakers. Best regards, -- Al P.S: @Marqmike, thanks very much for the nice words in your post yesterday. |
@Daveyf, what model are the speakers and what model is the sub? Also... I also have a REL sub hooked up to the amp, it is left, right and ground at the amp, as both the amp and the REL are balanced designs.
I presume you are using a three-wire cable to connect the sub’s Speakon connector to the outputs of the amp. Does "ground at the amp" mean that the ground (black) wire in the cable is connected to the amp’s chassis, or to a terminal that may be provided on the amp as a means of connecting to its circuit ground, or does it mean that the ground wire is connected to one of the negative output terminals of the balanced amp? And if the latter, is the ground wire connected to the negative terminal of the channel that produced sound when you misconnected the wires, or to the negative terminal of the channel that didn’t produce sound? Regards, -- Al |
@Daveyf, thanks for providing the additional info. Sounds good re the connections of the REL sub. You have NOT made the mistake I’ve seen more than a few members here describe having made, in which the ground wires of REL subs have been connected to a negative output terminal of an amp having balanced or bridged outputs. Which depending on the internal grounding configuration of the sub and the amp may often work ok, but depending on those factors risks the possibility of hum, sonic degradation, or even damage to the sub or the amp. Given that, I think we can rule out the presence of the sub as contributing to the consequences of the miswire at the speaker terminals. Also, while I couldn’t find a manual for the Sonus Faber Guarneri Homage speakers, I found a couple of indications that they are suitable for biamping, including this statement by no less than Martin Colloms in a 1994 review in Stereophile: The filters are nominally 6dB/octave over the crossover range, augmented by additional components to shape the acoustic output. The treble high-pass section thus has three elements: two film capacitors and an air-core shunt inductor. For the woofer’s low-pass section, the primary element is a large series air-core inductor with an RC Zobel network and an additional film capacitor. The multi-way binding posts allow for normal and bi-wiring, or even bi-amping.
I would conclude from this and from user comments I found elsewhere that it is a near certainty that the high and low frequency sections of the speaker are not interconnected in any way. Given that, I don’t see how the miswire you described could have resulted in damage to anything. I would have expected the result to be that both speakers would have played, but with poor sonics as a result of the high and low frequency sections being driven with opposite polarity signals. I can’t explain at this point why one speaker would have produced no sound, assuming there weren’t any loose connections. But perhaps Ralph or Roger will have some further thoughts as a result of the additional information. In saying this, btw, I’m interpreting your statement that the misconnection at the speakers that involved "one positive cable on positive, one negative cable on positive, one negative cable on negative, one positive cable on negative" did NOT mean that the positive amp output was connected to both + and - of the SAME section of the speaker, and did NOT mean that the negative amp output was connected to both + and - of the SAME section of the speaker. In that situation no sound would have been heard at all, from any speaker or speakers that would have been connected that way. Although again, even in that case no damage would have resulted to anything. Best regards, -- Al |
Ramtubes 12-26-2018
daveyf:
first the swapped polarity from amp to speaker does no harm and most will not hear any difference as long as the swap is the same on both sides. All you have done is invert absolute phase which has been a point of contention for many years.
As to one speaker not playing. probably just a loose wire.
Roger, see my lengthy response to Davey dated 12-22-2018, as well as Davey’s original statement of the question on the previous day. I believe that he did not simply invert absolute phase, but instead he caused the high and low frequency sections of the speaker to be driven with opposite polarities. On the speaker that was producing sound, that is. That would of course have adverse sonic consequences, while not causing any damage. On the speaker that was not producing sound a loose connection is one possibility, but another possibility (which I cited in my post) is that he was applying the + output of the amp to both the + and - terminals of one section of that speaker, and the - output of the amp to both the + and - terminals of the other section of that speaker. Which of course would result in no sound (and no damage), since no voltage difference would be present between the + and - terminals of each section of the speaker. Regards, -- Al |
@Jafox, to add to the responses Roger and Ralph have provided to your question, fyi I believe that as a special order item Cardas can supply XLR-female to RCA-male adapters that leave pin 3 unconnected, rather than shorting it to ground (pin 1) as is done by most such adapters.
Regards,
-- Al
|
@ffzz,
Yes it is. See two of my posts ealier in this thread, dated 11-26-2018 and 11-27-2018, which appear on pages 5 and 6 of the thread if you have it sorted with the oldest posts first.
Regards,
-- Al
|
@ffzz, you’re welcome! My former VAC Renaissance 70/70 MkIII amp is a class A amp employing four 300B power tubes per channel, in a push-pull parallel configuration, and is rated at 70 watts per channel. It is a 100+ pound beast, which I believe consumes something like 700 watts of AC at all times, converting most of that power into heat that is injected into the room. The Pass XA25, as you realize, is rated at 25 watts into 8 ohms and 50 watts into 4 ohms, operating in class A, and it is specified as consuming 240 watts of AC. John Atkinson’s measurements that were provided in conjunction with Stereophile’s review, though, indicated that it is capable of providing 80 and 130 watts into those impedances, respectively. He stated that some of that disparity is due to the fact that Pass bases its power ratings for the amp on much lower distortion percentages than JA uses, and presumably a lot of that disparity reflects the XA25 transitioning to class AB when outputting more than a certain amount of power. As I mentioned in my earlier posts that you saw my speakers are nominally 6 ohms, and have an unusually flat impedance curve as well as relatively high sensitivity, which makes them very versatile with respect to amplifier selection. Since like most solid state amps the XA25 is designed to provide an essentially constant output voltage into varying load impedances, for a given input voltage, (as long as the amp is operated within the limits of its maximum voltage, current, power, and thermal capabilities), even if we assume the very conservative 25 and 50 watt numbers it can be calculated that it is capable of providing at least 33.3 watts into 6 ohms. I wouldn’t say that the XA25 has "an edge over the VAC amp." They are both wonderful amps, in their own ways, and some non-sonic reasons factored into my decision to change. See the comment I left two days ago near the end of the following thread, as well as a subsequent comment by member "1markr": https://forum.audiogon.com/discussions/pass-labs-xa25-amp-and-bw-804-d3Best regards, -- Al |
@flashbazbo, the caveats that occur to me regarding adding a BNC output to your player are:
1) Be sure to select a 75 ohm BNC connector, rather than a 50 ohm BNC connector.
2) Looking at rear panel photos of the player, as might be expected it appears that the ground shell of the RCA jack is isolated from the metal panel. So you would want to select a BNC connector that is similarly isolated. Such connectors are readily available, but many BNCs are designed such that their ground shell is connected to chassis.
3) Waveform quality might benefit, at least slightly, if you were to disconnect the internal connection to the RCA jack and connect it directly to the BNC, rather than jumpering between the two connectors. You would not want to use both connectors simultaneously anyway, as significant impedance mismatches would result if 75 ohm loads were applied to both outputs. And if a need ever arose to connect an RCA plug to that output, BNC-to-RCA adapters are readily available.
Regards,
-- Al
|
Regarding differences in cable propagation velocity as a function of frequency, the following paper (which I and another member had referenced in posts here a few years ago) appears to me to be credible as well as informative: http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf See Figure 2 of that paper, although it addresses a coaxial cable rather than speaker cables. It can be seen that propagation velocity does decrease considerably at low audio frequencies compared to high audio frequencies. However, even at 20 Hz the propagation velocity, while much slower than at higher frequencies, is still about 5,000,000 meters per second, easily fast enough to be utterly inconsequential in the context of a home audio system, despite claims in some marketing literature to the contrary. As Ralph aptly said in another context here not long ago, where there is an effect there is snake oil for it. I would add that is particularly likely to be true when the claimed effect is not or cannot be looked at in a quantitative manner. ... the velocity of propagation of the signal (versus the velocity of the actual electrons) is determined by the dielectric or insulation material that the electromagnetic wave is predominantly traveling through. I believe that this statement is correct, and is unrelated to skin effect. Cable propagation velocities are usually somewhere between around 50% and 95% or so of the speed of light in a vacuum, and are dependent on the dielectric constant of the insulating material surrounding the conductors. Numerous references can be found on the web in support of that. The reason is that signal energy is conveyed in the form of an electromagnetic wave (rather than by the associated but vastly slower "drift velocity" of electrons), and for the most part that wave propagates outside of the conductors, within the dielectric (aside from a small fraction of that energy that is absorbed by the resistance of the cable itself). Again, though, whether an audio signal propagates from one end of an audio cable to another at 1 nanosecond per foot (close to the speed of light in a vacuum) or at 2 nanoseconds per foot, or somewhere in between, is utterly inconsequential. And if a 1 ns/foot cable sounds different than a 2 ns/foot cable, the reason is something else. Regards, -- Al |
I cannot recall any very detailed discussion of precisely what very good to excellent to superb implementation [of a DAC] actually involves.
Yes, there are general references to the quality of the power supply and especially to the quality of the analogue output stage; but nothing (that I can recollect, anyway) that goes deeply into the details.... IMO a major reason that the discussions you referred to have not delved very deeply into the details is simply that the details that are involved in the design of a high quality DAC, and the opportunities for the designer to overlook subtle issues that can adversely affect performance, are so vast in number that it would be impractical to address them in anything resembling a comprehensive manner. And it would be misleading to single out just a few of those details for discussion, while overlooking countless others. That is of course true to some extent in any sophisticated electronic design, but it is especially true in the design of a component that encompasses high speed digital circuitry, D/A converter circuitry, and analog circuitry all in close proximity. One major variable that usually seems to be overlooked in such discussions is the criticality of the design of the printed circuit board itself, including where the chips are placed, how signals are routed within the board, and how power and analog and digital grounds are distributed and "decoupled" (loose translation: "kept pure"). Take a quick look at the Table of Contents of the book "High Speed Digital Design: A Handbook of Black Magic," written by a noted authority and consultant on that subject, and at some of the pdf’s linked to in the "Downloads" section near the bottom of the latter page. You’ll get a small idea of the complexities that can be involved in the design of purely digital high speed circuits. Add D/A converter circuits and analog circuits into the mix and the opportunities for a design to become less than optimal grow dramatically. As the saying goes, the devil is in the details. And IMO what usually accounts for much and perhaps most of the difference between very good and excellent and superb implementation is simply the knowledge, expertise, and experience of the designer. Regards, -- Al |
|
