Schiit Vali 3 – build
Regarding the electronics design of this one small device from Schiit, this is where things get really interesting. First off, this is obviously a hybrid design, and even more, it’s not that usually found tube and FET contraption, because in the stage after the tubes it actually used BJT (bipolar) transistors that actually do the heavy lifting (which means providing the current for your headphones). This is not that quite often found configuration, because many manufacturers tend to use tubes with that field-effect transistors, since they market the latter as having tube-like characteristics, which is by itself quite, well, debatable. I asked Jason Stoddard, one of the founders of Schiit, to explain such a peculiar hybrid design and the reasoning behind it. And, as usually, I got an answer that can be seen as more than elaborate:
“Oh boy, this is a good multi-part question that gets to the heart of the matter. Let’s break it down a bit.
First, yes, Vali 3 is a hybrid, which means it uses both tubes (in this case, a single 6N3P dual triode tube, one triode for each channel, and a BJT, or bipolar junction transistor, output stage). This combo, tube and BJT, is a bit odd. A lot of people have gotten the idea that tube and MOSFET is the way to go, because of the oversimplification “MOSFETs are more like tubes.” Actually, they aren’t. And both BJTs and MOSFETs have their advantages. But to keep this on point, yes, Vali 3 is a hybrid, using a tube for voltage gain and transistors (BJTs) for current gain.
Second, advantages. Absolutely. The dream of a hybrid is to use tubes for what they are good for (voltage gain) and transistors for what they are good for (current gain). A great hybrid should deliver the best of both worlds–a low-feedback, linear amp that can deliver much more power than the tube can by itself. Of course, a non-great hybrid might deliver the worst of both worlds, so there’s always that challenge. And the biggest advantage of a hybrid is in the elimination of an output coupling transformer for the tube, which is the way you usually get higher power out of a tube into lower-impedance loads. However, the transformer is expensive and has its own nonlinearities. An output transformer wouldn’t fit the size or budget of Vali 3. So, a hybrid allows us to do smaller, more efficient, less expensive products.”
One thing that we should actually take note here is that lack of the coupling capacitors that usually many tube amps have. This usually leads to a clearer and more direct sounding. Obviously this will depend on many factors, and especially on a certain design choices in a certain amplifying circuit, yet the fact that there’s no coupling cap is always a good factor. Then there’s that thing that the tube here doesn’t actually act as a buffer, but it in reality amplifies the signal. This is what Jason has to say on this one thing:
“Yes, in the case of Vali 3, the tube provides all the voltage gain. Which means you’re actually getting a tube-like distortion characteristic, and a tube-like soft clipping behavior. Which should help Vali 3 to sound “more like a tube.” The catch with tubes is what do they really sound like? The tube alone (but maybe with coupling capacitors)? The tube plus output transformer? The tube plus hybrid output stage? With Vali 3, we were shooting for something that credibly delivered some of the “midrange magic” of tubes, and using the tube for voltage gain was a key part of how we got there.”
And then, we have that oh-so-bold statement about the plate voltage and the quote from the Schiit’s website that “tubes love voltage”. Yes, they do:
“Unlike semiconductors, tubes have no actual physical connection between its amplifying and control terminals. As in, the anode, cathode, and grid all terminate in a vacuum. Tubes work by heating the cathode (via the heater, one reason tubes run so warm) to the point where an electron cloud forms over the cathode. That cloud then is attracted to the anode by the voltage across the tube, causing current to flow. The current flow is then modulated by the grid. But the upshot is, the less voltage, the less current flow. The less voltage, the less chance you have to keep the tube in its linear operating region. So, in general, more is better. Of course, there’s an end to the party–tubes typically have a maximum anode potential specified. In the case of the 6N3P tubes we use, 100V is a good anode voltage.”