Apart from that – the amp looks great. And is just another example of the clever approach from the guys at Schiit. By the way, as I mentioned – and as the name suggests – this is the second version of this power amp. This is what one of the Schiit founders, Jason Stoddard can tell us about the whole idea:
The original version of Aegir (20W into 8 ohms, a bit weaker than Aegir 2) came about as a “what-if” thing, most likely at the local brewpub. As in, “what if you took a Vidar and ran it in Class A? What kind of power would you get out of it? Turns out the answer was about 10W. The thing is, those 10W sounded pretty good. And when I came up with the whole “Continuity” idea, which extends the benefits of Class A outside the Class A bias region, that became a much nicer 20W. Still weak by the standard of a 500W Class D thing the size of a pack of Marlboros (slight exaggeration), but plenty to run efficient speakers. And, with everyone concentrating on bigger power amps—lots of 100+W entrants out there—I figured it might be a good idea to go the other way, and do a low-power, but high-quality Class A-ish amp instead. Turns out it wasn’t a bad idea. Aegir did pretty well in terms of sales.
And while we’re at it, he also explains how does the Aegir (2) looks when we consider that there’s a Vidar (price-wise) slightly beneath it and Tyr way above:
Vidar 2: do-all. I recommend most people start with one Vidar 2 if they have typical or difficult speakers. One Vidar absolutely lights up Magnepans.
Aegir 2: for more efficient speakers. Loves Zus and Klipsch, but horns are not a requirement—just naming a couple we have around.
Tyr: if you have the room, the end-game. Doesn’t care too much about the speaker load. Differential. Completely bonkers.
As you can see, Schiit’s very technically-savvy Jason doesn’t leave us with any room for confusion – it’s the Tyr that is still the company’s best power amp for now. Well, for that price it should be. This doesn’t at the same time means that the Aegir 2 is less interesting, because there’s actually a lot of going on under the hood. When I asked Jason whether this is a “discrete, solid-state amp with bipolar output stage” he got some interesting things to say:
Yep, with a couple of our typical tricks, and a couple of our less typical tricks. The typical tricks are:
- Aegir 2, like most of our amps, is discrete, current feedback, low loop feedback, and bipolar all the way through. Current feedback amps, or CFAs, are still a bit weird in a world where most Class A or AB amps are Lin/Blameless voltage feedback, high-loop-gain designs, but it’s one we prefer for sonic and transient reasons.
- Also like most of our amps, it has a completely linear power supply, no switchers, just the good old big-transformer and big-filter-caps thing you’ve seen for decades. That’s why it weighs like a brick.
The less typical tricks are:
- Continuity™, which is our way of extending the benefits of Class A bias outside the Class A bias region, or, in the parlance of Bob Cordell and John Broskie, a non-gm-doubling, or non-transconductance-doubling, design. Transconductance doubling is where the output stage has higher transconductance in the Class A bias region, since both N and P devices are conducting, and lower transconductance once outside of the overlap region. Continuity adds transistors that come online when the amp exits Class A, making the transconductance more constant. More below on this, since you ask specifically about the type of Continuity we are using here.
- Halo™. I’ll tackle that below as well
As you can see, there’s a lot of going on here, and there are things that deserve even more explanation, like Continuity and Halo. Let’s start with the former, which is also directly connected to the fact that the Schiit uses both the NPN and PNP transistors on both rails of the amplifier, which is quite unusual. This can be seen as really important, especially when we consider that’s it’s the final outcome that counts here. And this is how Jason explains to us, why:
One of the biggest benefits of Continuity is that it runs both NPN and PNP transistors on both sides of the rail at all times—which is why it potentially could be better than Class A. NPN and PNP transistors, even “complementary pairs,” even “matched complementary pairs,” are never really matched nor complementary, because physics. One look at the beta behavior vs output current and temperature for claimed “complementary pairs” will put that idea to rest. So, since you can’t fight physics, the best way to achieve balance is to run both types of devices all the time. Then they are complementary. That’s one of the things that makes Continuity special. (Oh and no, Continuity is not a CFP output stage, which has gain—it has no gain.)