Tube Preamp Reviews

Removing the top panel reveals the major surgery your PAS will have undergone in Minnesota. The original mains transformer is retained, but has been moved outside the chassis, to its right rear. In the space it once occupied, AVA has installed a new power-supply board featuring a nest of modern electrolytic filter caps. The heater winding from the transformer now feeds a twin silicon-diode rectifier, followed by no less than 44,000µF of smoothing, the heaters being DC-powered to avoid hum. (The four tubes must be identical for correct heater voltages.) The original 12X4 rectifier tube is retained for the B+ supply, but now feeds first a capacitor-resistor-capacitor pi filter, then four individual dropper-resistor/330µF reservoir cap combinations, one each for both channels of the phono board and both channels of the line-stage board. HT voltages, though unregulated, are thus highly smoothed, and are 182V and 236V respectively.

The new jackfield is on the left rear of the chassis; though the sockets appear very close together, there is just enough room to use Monster M1000 with its large-barreled plugs. The line-level inputs other than "Tape" are taken by individual left and right wiring looms to the ceramic-wafer selector switch. (This shorts unselected inputs to ground to minimize crosstalk between inputs.) The phono input signals are taken by twisted pairs to the phono board, the outputs from which also go to the selector switch.

Standard phono input impedance is set at 47k ohms, with parallel capacitance kept as low as possible, though purchasers can specify other resistance values if they so desire. It would also seem an easy task to add input capacitance, but Frank Van Alstine feels that electrically slugging a moving-magnet cartridge's mechanical resonance peak to give a supposedly flat steady-state response is fundamentally unsound. The LF pole of the RIAA equalization is passive, the two-triode circuit essentially running open-loop at low frequencies, and a complete 5751 twin triode is used per channel to maximize channel separation.

The input selector switch keeps the channels physically separated by around 1", also to minimize crosstalk, and is followed first by the source/monitor switch, then by the stereo width and balance controls. The signals are then routed to the volume control and from there to the line-stage board, which features a second pair of 5751 tubes. As there is no active stage between the line-level inputs and this control, there is effectively an infinite overload margin on these inputs (or at least until the pot melts).

The construction is to a high standard, with a neat, rational layout of wiring and high-quality parts used throughout, including metal-film resistors, and polystyrene and silver-mica capacitors in the RIAA equalization. Having a traditional single-ended power supply, the circuitry must be AC-coupled, and a total of four DC-blocking series capacitors have to be used in each channel's signal path. These are matched, high-quality, plastic-film types from Panasonic.

Sound Quality
Whether via the transformer or straight in, the sound of the phono input was lightish in balance, with a slightly forward upper midband and a modest degree of exaggeration at the very top of the audio band which added a slight "tinkle," a wispy edge, to cymbals and snare drum. The midrange appeared a little "thinner" compared with the original on bypass tests. The degree of such tonal modifications was mild, however, and certainly not objectionable.

Where the Super-PAS scored was the clarity with which it presented musical detail, both via the phono input and via the high-level inputs. It doesn't achieve this by exaggerating or spotlighting detail—some components achieve this by a presentation which I once compared to reading a book with a spotlight shining over your shoulder—but by allowing you to "reach" into the sound, almost to examine each component of that sound on its own as you could in real life. On the Gluck track on the HFN/RR test CD, for example, the tonal and imaging differences between the half recorded with the Calrec Soundfield mic and that captured with the pair of STC ribbons were superbly revealed, as were small details of each individual sound—for instance, the way in which the dome over the stage had modified the tonal quality of the piano to a greater extent than the flute. (Though when the door shuts at the end of the Soundfield version, it still needs better resolving power than the Super-PAS's to reveal that it was Hi-Fi Answers Editor Keith Howard who had closed it.)

Low frequencies were lighter in weight than my reference in this department, the Krell KRS2, which has more "authority," but bass extension was excellent, bass percussion instruments having the combination of clean leading edge and powerful following tone that they do in real life. Midbass definition, however, was not as good as I would have liked, bass guitars and double bass being a little woolly, sometimes acquiring a "grumbly" quality, where the body of the tone seemed detached from the supporting transient framework.

The stability of the imaging was excellent for what is basically a budget-priced component. Laterally, the soundstage presented by the Super-PAS was very well-defined and stable, but image depth via LP was not up to the standard set by the Audio Research SP10 or Mark Levinson No.26, the walls of the hall on my piano recording on the HFN/RR Test CD being less well-defined. In addition, as well as there being less depth apparent, that depth was, for whatever reason, more obvious in the center of the stage than at the edges, giving a somewhat triangular impression. But, come on, look at the difference in price between the Super-PAS and those thoroughbreds!

Auditioning the line inputs revealed a signature similar to the phono input, though the low frequencies now had a little more weight. Overall, I enjoyed the time I had with this preamp, feeling that its sonic performance was excellent at this price, and equal to that of models costing up to $2000.

Finally, in his covering letter, Frank Van Alstine suggested turning the Super-PAS 3 on and off while playing music "to note how long it continues to play, and how nicely it shuts down." It took me some time to pluck up the courage to do this. I have come to love the Celestion SL700s, and have had experience of the humongous, speaker-blasting thumps some solid-state preamps produce when turned off while still connected to a functioning power amplifier. Well, turning off the Super-PAS gave just the result Mr. Van Alstine had suggested: the music continued to play for about 20s, then faded to black. Turning it back on again just brought back the music, with nary a thump nor bump to alarm the faint of heart. The use of a tube rectifier, which will turn on and off in a controlled manner, must be at the root of this graceful behavior, and I assume that the decoupling of the reservoir caps from the tube diode with series resistors prevents it being overstressed by the B+ supply's current demands on turn-on.

Conclusion
I was impressed by Frank Van Alstine's reworking of the classic PAS design. The standards of engineering and construction are high, the lightish-balanced sound features great clarity in the midrange, low-frequency definition is relatively good, and the treble is refreshingly free from grain. More important, the preamp never fails to be musical. Recommended in Stereophile's Class C, the AVA Super-PAS Three is one of the least expensive, true high-end preamplifiers around, only losing out to competing designs like the PS Audio 4.6 in that it lacks a moving-coil stage.