Tube Power Amp Reviews

When I measured the Wavelength Cardinal, another single-ended tube amplifier, I commented on its poor test-bench results and noted that it might have been a hot design in 1940. It's only fair that I make the same comment about the Jadis. The Jadis also measures worse than the Wavelength in one very significant way; it has a severely compromised bass response. Though I haven't heard the Jadis horn loudspeaker, which JS used in his review, many commercial horn loudspeakers don't have a very extended bass. [Note JS's auditioning comments about the system's bass quality—"Punchy, differentiated, controlled, and of an excellent overall character, it never disappointed. At the lowest late-night listening levels, as mentioned, it could sound small and rather inconsequential...", which appear to correlate with the measured response.—Ed.]

And in our February 1996 issue (p.57), Sam Tellig commented in his column on the SE300B using the Infinity Compositions Prelude loudspeaker, which not only has a high sensitivity, but which also has its own built-in, powered subwoofer, which would eliminate the need for good bass response from the Jadis.

Like all single-ended tube amplifiers, the Jadis SE300B is a very specialized product, and must be carefully matched to its partnering loudspeaker. It's not a general-purpose amplifier. As a reviewer assessing a wide range of loudspeakers, for example, I simply couldn't tolerate its low power and load-dependent frequency response. Nor would a loudspeaker that sounds its best on this amplifier necessarily make a good all-around reference. Make no mistake, the Jadis (as with many other tube amplifiers, particularly single-ended ones) is a tone control. A crooked wire with gain. This is not a reflection on Jadis's implementation—except in the bass response. It's simply the nature of the beast.

The measured frequency response with our simulated loudspeaker load, at least above 100Hz, is not inconsistent with the appeal of such an amplifier. The small suck-out at 180Hz will tend to open up the sound, though to a limited degree the amplifier's LF distortion will help fill in the hole, substituting distortion—however pleasant-sounding—for real information. The rise at 1–2kHz will give the sound a bit of immediacy and palpability; the dip at 4–5kHz will soften recordings that are edgy in the mid-treble and sweeten the sound. And since I suspect most potential customers for a design such as this are heavily into analog playback, the HF rolloff will help soften the HF rise typical of moving-coil cartridges.

To address the whole Pandora's box of measurement vs listening, I believe the listening results are primary. I also believe there are certain boundaries within which a product's performance must fall to make it recommendable, even if it sounds "good." Otherwise, we're dealing with trying to match complementary colorations, the minefield of audio never-never land. I won't attempt to specify what those measured boundaries should be—yet another minefield. But, in my opinion, the Jadis SE300B clearly falls outside them. We've said it before, but it bears endless repetition: does it sound "good" because of the way it measures, or in spite of it? You can argue that we haven't yet learned to measure all the right things, and I'd be first to agree with you. But high static-distortion figures, frequency-response aberrations, and lack of power are important considerations, and will be clearly audible in many, perhaps most, applications.

Absolute accuracy is impossible, and beyond a certain point we're forced to fall back on what sounds "good" to us—what most closely results in a believable simulation of the real thing—rather than a literal translation, which we cannot yet accomplish, of what's on the recording (never mind the accurate recording of the source itself, an even bigger problem). It's the balancing of such tradeoffs that makes audio such an endlessly fascinating pursuit. Short of the immediate perfecting of everything in the record/playback chain, I wouldn't have it any other way. And there's also that large, fuzzy gray area where the degree of achievable—and achieved—accuracy will continue to be debated. But outside of this area, how far are we prepared to go? Should we throw away some of that objective accuracy we know we can achieve for "sounds good?" That is the crux of the problem.

But as I said about the Wavelength Cardinal, if you fall in love with the sound of the Jadis in your system, and have the disposable cash, buy and enjoy. But make the decision by consciously weighing the options—knowing what you're gaining and what you're throwing away in the process—and not because of the current drumbeat for single-ended tube amps as Magic Feathers. They're not.—Thomas J. Norton