Aikido Phono 1+ MM preamp review

[JoshK]

Doubly odd as Aikido has for decades been associated with a tube circuit from Glassware.

I knew it as John Broskie's Tubecad.com blog (Glassware came about iirc when he started selling boards and then kits to advid diyers, like my former self). I've been out of the tube DIY scene for at least a decade (lack of time and space), but read his blog regularly since the early years. Broskie was well aware of the martial art Aikido and purposefully used the name as a metaphor to explain the concept of mains noise rejection by injection and played with the concept from Aikido the martial art. This is from memory of a blog probably 15 years ago at minimum.

Then it explained it wasn't simply a concept for triodes and could be used with 3 JFets (he also explained why you wouldn't use 4, like you would with triodes). Then later he did the Aikido Phono, with the RIAA circuit in between 2 Aikido circuits. I don't recall if he ever shared a full Jfet version of the phono concept but from the description of this circuit, it sounds like this maker used that idea.

 

[Grotti]

In the description on eBay the designer mentions to use 2SK117 FETs. Since I might know something about loudspeaker design I am a layman to active electronic parts. Can anyone share some wisdom about those transistors? Not that it matters much....

What was new to me is Otto's statement, that the output stage works in Class A .... But it doesn't seem to have an impact on the measurements (for good or bad). Or does it?

 

[pma]

In the description on eBay the designer mentions to use 2SK117 FETs. Since I might know something about loudspeaker design I am a layman to active electronic parts. Can anyone share some wisdom about those transistors? Not that it matters much....

What was new to me is Otto's statement, that the output stage works in Class A .... But it doesn't seem to have an impact on the measurements (for good or bad). Or does it?

2SK117 is a JFET with lower transimpedance than popular 2SK170, with fairly low noise.

https://www.futurlec.com/Datasheet/Transistor/2SK117.pdf

Otto Aikido design uses 2 class A Borbely followers/buffers, 1st behind the linear gain input stage to drive passive RC RIAA equalization, the 2nd one at the output. Yes they work in class A (that circuit structure can do nothing else than class A, because it is not a push-pull, but SE). It has an impact on measurements, as you can see only H2 and H3 in spectrum in my post #1. It is free of higher order harmonics. But, as I repeated several times, it must be loaded with quite high load impedance to have such spectrum.

To understand what is Borbely buffer, please read

JFETs: The New Frontier, Part 2

The basic circuits on which Erno Borbely based most of his projects are described in “JFETs: The New Frontiers, Part I and II,” published in Audio Electronics AE 5/99 and 6/99, and “All-FET Line Amp,” published in audioXpress May 2002. As the author admits, these were the “building blocks” on...

audioxpress.com

and watch the images 15C and 16A.

I think I have quite exact circuit schematics of this phono preamp. I can send it to you in a private mail, if you wish.

 

[JoshK]

My circuit understanding comes from working with tubes, even though my main reference system uses none. It was a hobby. They sound great easily, whereas a lot of SS needs a lot more understanding and real engineering to do their best. My non-expert opinion. I think it has to do with the inherent linearity of triodes, predominance of H2 and H3 distortion which our ear masks and soft clipping, aka compression rather than nasty clipping.

I am not one to hold tubes over the best SS designs, but they are easy to make sound good even for a hobbyist. At worst, they sound euphonic, which is a musically complementary.

IIRC, the first Jfet works like a triode in gain mode. These curves looks like triode curves. The next two are stacked as followers, although in SS speak, they might use different nomenclature.

Class A circuits are hard to mess up if the operating points are chosen correctly. They are just inefficient. For preamps, we don't really need efficiency as a high priority. We want linearity. However, distortion early in the chain causes amplification (of said said distortions) and dissonance products (distortion of distortion) downstream.