Here are Stereophile's output impedance measurements for the SP16L line-stage preamp, for which the ARC specs are identical to what Mapman quoted: http://www.stereophile.com/tubepoweramps/901/index6.htmlThe measurements indicate 262 ohms at mid and hi frequencies, and 549 ohms at 20Hz. Into a 600 ohm load (sic), they measured a loss at 20Hz of only 2.5db (see Figure 1). By my calculations, into a 10K load the 549 ohm output impedance would result in a loss at 20Hz of less than 0.4db. So the output circuit, and particularly the output coupling capacitor, of the SP16 appears to be designed more conservatively than in the case of most tube preamps. And the recommendation of a 20K minimum load appears to be conservative as well. Regards, -- Al |
Amen from me too, of course.
I found the review very quickly because I was too dumb to think of looking in the component category listings. It is in the Power Amp category, because it was reviewed in conjunction with an ARC VS110 power amp. I just entered "Audio Research SP16" into their search function.
Best regards,
-- Al |
Newbee (& Bob), thanks very much for the kind comments; much appreciated. I've learned a good deal since I've been here from your posts, too, as well as those of all of the other regulars who've participated in this thread.
My statement of "less than 0.4db" loss at 20Hz with a 10K load was a quick rough swag, and the actual number is probably close to 0.1db (and almost certainly less than 0.2db). The exact calculation is a little complex, because the preamp's output impedance is partly capacitive and partly resistive, and resistive and capacitive impedances don't sum linearly.
But with the numbers being that small at 20Hz, and undoubtedly totally negligible at say 30Hz and above, there's no need to be more precise. 0.1db is the commonly accepted rule-of-thumb for conservative level-matching, but that figures to be driven mainly by mid-range frequencies, where the ear is vastly more sensitive.
Thanks very much again for the nice words.
Best regards,
-- Al |
I'm currently using a Burson AB-160 buffer between my McCormack DNA-500 amp and VTL preamp. I'm waiting for it and associated cables to break in and sonics to stabilize. I'll report back with developments after it gets some significant hours on it. Emerson, as I recall it was not readily possible to determine the low frequency output impedance rise of your VTL 2.5. However, Stereophile has numbers on the 5.5, which it describes as being descended from the 2.5, and the two models have nominal output impedances which are very close to one another (195 and 200 ohms). For the 5.5 they measured 184 ohms at mid/hi frequencies, and 409 ohms at 20Hz, which is a very minimal rise that would have negligible frequency response impact into 10K. They also have numbers on a couple of other VTL models, the 6.5 (150 ohms at 20 Hz for the unbalanced outputs) and 7.5 (69 ohms at 20 Hz for the unbalanced outputs). So it would seem that VTL's design approaches reflect sensitivity to this issue. Regards, -- Al |
Glai, it's only possible to offer speculation in response to your question, without having more complete information on the particular designs (which may be unobtainable). But I wouldn't be surprised if a tube preamp with a nominal output impedance spec of 600 ohms were to rise to several thousand ohms at deep bass frequencies.
And the 600 ohm impedance itself could have consequences at high frequencies as well, if your interconnect cables were long and/or had high capacitance per unit length. That could very conceivably account for the degraded transient attack you mentioned, by rolling off the upper treble.
Also, perhaps in the case of the particular designs you were using, there was some degree of correlation between lower output impedance and better quality parts in the output stage. As well as the possibility that variables elsewhere in the designs, unrelated to the output stage and impedance matching effects, might have accounted for the differences in those particular cases.
Regards,
-- Al |
Glai -- Wow! I looked at your system description, and I must say it is utterly magnificent. Congratulations!
As to why the Ref 3 did not give you the tight bass and the better transient response that the other two preamps do, I have no ideas to offer. What I can say is that I doubt it has anything to do with the somewhat higher output impedance of the Ref 3, considering that your interconnects are short, your power amp input impedance is reasonably high, and considering its outstanding reputation among many of our members, and the excellent reviews of it that I recall seeing.
Emerson -- Glad things seem to be shaping up. Again, there is no technical explanation that I can envision for the effects the Burson buffer stage seems to be having, and I doubt that it has any relation to the impedance issue this thread has focused on. As we all know very well lots of things in audio are both unexplainable and unpredictable.
Regards,
-- Al |
Guido -- While I certainly don't question your subjective comments, I don't think that the statement that 45 amps from a balanced amp output is roughly equivalent to 90 amps from unbalanced amp outputs is correct. If the spec is 45 amps (and btw I don't see that spec at the Bel Canto site), a maximum of 45 amps (plus whatever margin is built into the specification) would flow between the amp's red speaker terminal and black speaker terminal, regardless of whether both of those terminals are actively driven, or only one of them is actively driven.
The specs I do see are 500W into 8 ohms, 1000W into 4 ohms, and a rated minimum load of 2 ohms. If we assume that 2000W can be delivered into 2 ohms, that corresponds to a current of 31.6 amps.
Mapman -- Congratulations on the new amps! Enjoy!
Darkmobius -- As you realize, Class D technology is at the opposite end of the spectrum, relative to Class A, in terms of the amount of power that can be delivered relative to the amp's size, weight, and heat generation. Your characterization of them as "voltage amplifiers," though, is correct in the sense that they typically have extremely low output impedances (or, equivalently, very high damping factors). That allows them to deliver an output voltage which is directly proportional to input voltage regardless of load impedance, within the limits of their current capability (which in this case is very substantial).
Regards,
-- Al |
|
