What I found amusing about the review was JA's inability to use his normal measurement techniques due to the micoprocessor controlled bias (or something. I don't know how this stuff works.)
It's like Jason went out of his way to build an amp that would poke a finger in the eye of the measurement crowd. Note that I don't mean he actually did, just that I felt that way about JA's tribulations.
I think measurements need to catch up to the microprocessor inhabited world we live in now.
I actually didn't go out of my way to make it difficult to measure, but, on reflection, I can see how it seems that way.
Kinda cross-posting from another thread (edited for this thread), a "cliffs notes" version of the Ragnarok whys/wherefores:
Why did we make the Ragnarok an “intelligently managed” amplifier, that continuously monitors and sets bias (as well as monitoring for faults and correcting for DC offset?) The short answer is to provide a more stable operational point (which should eliminate much of the “warm up” variability, at least IMO) and to eliminate the need for coupling caps or DC servo (which, to me, are both imperfect solutions, but we can have loooooonnng discussions about that.) The long answer is in Chapter 27 where I discuss all the decisions that went into the development of Ragnarok, over the course of several years (but, if you have read this chapter, you know that there's a ton of rambling there beyond the microprocessor control.
So for what reason did it measure so oddly in Stereophile’s testing? First, de-biasing on continuous sine waves is part of Ragnarok’s operational algorithm. On extremely high volume pure sine waves, such as the 20 watt (1/3 power) one used in testing, it will eventually step the bias down to zero. A 1W pure sine wave won’t do a thing—it is below Ragnarok’s Class A bias. We ran hundreds of hours of music through Ragnarok in developing this algorithm to ensure it was not triggered by music, while watching the output of the bias ADCs on a screen. And, as Stereophile noted, Ragnarok does not de-bias with musical input. Again, we can have loooooonnnng discussions about whether this is “right.” (It does preclude the typical 1/3 power preconditioning test, for example.)
And the mystery? The mystery is why Stereophile could not reach our numbers for power output and THD. This is something that's difficult to guess about, since I wasn’t in the Stereophile lab to see how they were testing the product. As I have noted in 2015, Chapter 12, getting good, repeatable measurements is by no means a certainty. Was the test system set up for single-ended output? (Therefore shorting the negative outputs together, which is no bueno for circlotrons.) Was there a single-ended instrument in the mix? (Clipping a scope probe ground to the negative output of an inherently balanced amp also can cause big problems.) But all that is speculation. That said, I will be the first to note that Ragnarok won’t deliver the numbers that, say, a high-feedback, multiple-gain-stage amp will, so if you’re looking for PPM levels of THD (0.000X%), it’s not the amp for you.
I try to do the best to chronicle
what we do—and
why we do it—in the most transparent manner, in this thread. (Some would say, “in
excruciating, painful detail.”) We
want you to know who we are, what kind of decisions we have to make, and why we do things. It’s up to you to decide if what we say resonates with you, and if our products meet your needs. I hope you do…but I also understand there will never, ever, ever be 100% consensus.
All the best,
Jason Stoddard
Co-Founder
Schiit Audio