This review page is supported in part by the sponsors whose ad banners are displayed below

Intermezzo I. After the Munich HighEnd show Sven Boenicke and Gino Colombo had dropped off their combined system—sans source and cables—which Marja & Henk had nominated Best of Show in their report. My review focus was to be Sven's new WE-20 speaker. The previously reviewed Colotube monos came along for proof of concept. Boenicke's newest model had been commissioned to give the Colombos a high-efficiency flagship speaker. So I had 8-watt pedigreed 300B SETs with direct-heated high-gain 20A drivers. I had a 4-driver 3-way with Supravox tweeter and widebander, 18sound 12-inch woofer, rear-firing ambiance tweeter and honest 95dB sensitivity. I had the 10-watt SIT-1. I had a perfect setting to explore an all-transistor path, an all-tube path and the more interesting hybrid mixes. Run amp direct, my usual Eximus DAC with analog volume played transistor preamp. My ModWright LS-100 with Psvane CV-181T for half a premium 6SN7 per channel represented a non-exotic tube pre. How would the SIT mirror a superior transformer-coupled zero-feedback valve SET of the same power rating? How would it diverge?


On this particular speaker, the best sound came—unexpectedly!—from transistors pure. With only a 2000Hz+ tweeter in direct line of sight (tested with a laser finder, the 40° upfiring midrange hit the ceiling directly above my listening seat), the sound was dominated by omni traits. Think far-field balance. The ambient field superimposes itself on direct sounds. The DAC-direct valve amps still exhibited insufficient speed, incisiveness and immediacy to cut through this pretty but somewhat diaphanous energetically subdued presentation. With them the sound was all about mellifluousness but no spunk. Preceding the 300B monos with my 6SN7 preamp only made things thicker and worse (it did add mass in the bass but no greater control).


Replacing the Colotube monos with the SIT1 eliminated noise. The triodes were audible in front of the speaker drivers. The Vfets expanded bandwidth particularly on top, increased in-sight into the soundstage to benefit depth of field, exhibited better separation during complex passages, were texturally grippier, had better bass, were more finely resolved and more informative at low volume. If the triodes put oil in the water, the SITs were pure spring water. In this context the direct-heated power triodes had not one advantage though they were priced twice as much plus. Given the very happy combo of lightweight Berning 300B amp vs. Nelson's triode-curve transistors at Dan's, the apparent culprit in this encounter must have been the output transformer. Thus calling the SIT 'tube-like' fails if—and that would be the overwhelming majority—it referenced transformer-coupled valves. It does however become surprisingly accurate if one meant a single-ended OTL like the Berning. Then lucidity, speed and true directness step into the light to stand right beside the SIT (though the latter still played it gutsier and fuller than the 300B). Without such a qualification which really is an exclusion on a mass scale, fanciers of conventional SETs wouldn't really relate. With the SIT they'd expect a sedan with soft suspension. Instead they'd find themselves in a far sportier more exhilarating ride with greater feedback on road conditions.


Intermezzo II. "The white Tara has arrived. This you need to hear!" Naturally Dan's invite prompted another immediate visit. This Trafomatic commission—Dan knew the 10Y well and wanted to hear it in a preamp—took six month of R&D, "30% for the sound, 70% for the noise". The 10Y is very microphonic. It required heroic design measures to behave dead quiet in a linestage. The small power supply valves between its output bottles are two EZ81 rectifiers whilst voltage stabilization is handled by two each EL84 and EF86 pentodes. Because the 10Y's amplification factor of 8 was too low, designer Sasa Cokic put a 1:2 step-up transformer on the inputs to add passive gain. Unlike Maloney's Predactor, this eliminated a second stage and with it the dilution of the direct-heated power triode's sonics by another voltage-gain tube. Did this explain what followed?


With each SIT preceded by just one premium dead-quiet 10Y, now the sonic changes of the amp's bias adjustments had been stretched to profound. Forget any prior subtle. This approached most useful audio tuning feature ever designed. As a result the difference of 25%—halfway between full left and center on the meter's green window—was easily heard to become preference. Affected performance aspects included top-end clarity, transient definition and musical energy/charge (right = more), tonal fullness, bass heaviness, density and a mellower gestalt (left = more). Swapping the 91dB Aries Cerat 3-way Gladius for the 100dB single-driver Voxativ Ampeggio clearly changed the optimal needle position. What's more, the 10Y/SIT combo now made even better sound than I'd heard with the 71A on my prior visit. Dan had hit upon one of those mysterious magical 1+1=4 combinations. The operating point feature had turned into a powerful Photoshop equivalent for lighting, contrast and hue. Yet back in Villeneuve I couldn't repeat this with two transistor preamps or my 6SN7-based valve preamp. Now the bias changes had once again shrunk in scope to subtle where I'd started.


This raised an interesting question. Was a direct-heated power triode required to scale up the bias adjustment to the next power? And not just any triode but one not already harmonically mixed up with its own driver tube? Single-stage preamps with DHTs are very rare. Eddie Current's Balancing Act drives its 300B, 2A3, 45 or PX4 with a 6SN7. So does Manley Lab's NeoClassic 300B. Allnic's L5000 DHT uses a direct-heated driver tube. Supratek's preamps used octals. Only Coincident's Statement relies on two solo 101D, Wavelength's Royal on equivalent 71A and a few Experience Music models on the same types. As synchronicity would have it, Sasa Cokic announced himself. He wanted to hear his Tara in Dan's system. He offered to ship out his solitary 101D preamp sample which also runs its power triode pure as another single-stage design. Would this repeat the 10Y's 'boundary expanding' SIT action at Dan's? And would I be able to repeat it in Villeneuve?


Underlying rationale—or wishful thinking—fell back on known THD effects between driver/power tubes. In Sasa's opinion the impact a driver tube has on the output tube is seriously underrated. Much deliberate design work is required to trim the harmonic distortion interaction and precise 2nd/3rd-order balance one wants. The SIT was a de facto single-stage single-ended triode already. Stood it not to reason that it would only want a single 'driver' tube to precede it; and not simply a common 6SN7/12AU7/5687 type but a full-blown premium power triode with very specific THD fingerprint?


It seemed a possible if unexpected explanation. To progress from fluke to potential theory required at least a second such preamp and different source components to mix up context. As luck would have it, now our stage was set for just that.


Incidentally, moving in Dan's big 75TL Vilobhas confirmed the former OPT argument. Even the very best of purpose-wound output transformer tasked with significant impedance conversion for speaker consumption impacts circuit speed. Though to a rather smaller extent than the earlier Colotube/SIT comparison, the Vilobhas did trail the transistors in this regard. Whilst some might prefer the valves' slightly softer rounder handling of transients, it was an audible effect.

Sasa Cokic and Milorad Despotovic of Trafomatic Audio hearing the fruits of their labor

Sasa's Serbo-British referred to it as "a little too much sweetly" when referenced against the 10Y/SIT combo. Chatting with him afterwards about what we'd just heard, he fully agreed that with an amplifier, hearing 100% of any power triode is impossible. The high turns ratio of the output transformer to make 8 ohms or less throws away gain from the output tube. This requires another valve for voltage gain which alters the flavor of the power triode quite dramatically. To hear a pure triode mandates putting it in a preamp where an output impedance of 2000Ω remains acceptable to allow a single-stage circuit to produce sufficient gain. (An obvious fringe benefit as preamp is driving a fixed input impedance rather than reactive loudspeaker load). Here comes the shin kicker for unrepentant valve maniacs. Leaving such a pure triode maximally unmolested might require a single-stage amp to follow it. That's why Nelson's SIT1 might presently be the only amplifier to maximally exploit the aroma of a direct-heated triode. Truly.


To hint at the special magic of our 10Y/SIT1 encounter descriptively, envision a man living by himself perfectly self-reliant, content and complete. His inner feminine side is properly cultivated, nothing is missing. Yet it's not the same as suddenly having a real woman's touch manifest itself in his home and life. The SIT by itself was that perfect man. An ordinary preamp mated to it became an ordinary woman insisting on too many accommodations and changes. Give and take. Meddling. The 10Y preamp was the perfect womanly touch. No compromises. Completion of the already complete. All give, no take. Ideal yet not idealized. Very much real and attainable though admittedly rare if, as our theory had it, the full effect relied on this type of preamp. Time to return to Villeneuve and try my silver SITs with Sasa's one-up 101D preamp