A Letter form Poland

Pawel’s first circuit: a DC coupled headphone amplifier Download B2 Spice A/D circuit

17 June 2003

A letter from Poland (slightly rewritten/translated)

I have read all of your publications on the web and I urgently look forward to see any new issue of the Tube CAD Journal. After my long experience with tubes [10 years ;-) ], I am now trying to create audio projects based on solid state devices ‘cause tubes are almost perfect, but transistors aren’t — so it is much easier to build good sounding equipment with tubes — thus it is a bigger challenge to build something similar using only solid-state parts (well, for me, of course).

In all of my projects, I try to omit capacitors and I try to use only DC coupling without any global NFB. By the way, many times I have asked myself why someone as experienced as you hasn’t designed a good amplifier that fulfilled the above two demands? Even your bold designs often have global negative feedback...is this really necessary?

I attached a design of tube headphone amplifier that might be discussed by you in the next issue, if you like. The project may be changed by swapping out tubes for FETs and MOSFETs, which might bring bonuses in its ability to drive lower impedances, what do you think?

The second attachment is my attempt to approximate a design that fulfills the rules as stated above — sorry — it is solid-state (I know I’ll go to hell or at least to jail for this!). Could you comment on it too or give a general directions to improve it or exchange transistors with the tubes or make a great hybrid?

Pawel

Poland

The goal of “no capacitors, no feedback” is laudable—within reason. In designing my own equipment, I always strive to use the fewest capacitors and the shortest feedback loops possible or, preferably, no feedback—within reason. And reason tells me that if I use a plate resistor then I have added a feedback mechanism to my circuit. The same holds true if I use an unbypassed cathode resistor or even if I just use a triode, as each constitutes a feedback loop of sorts. The tip-off is “does distortion go down?” If it does, then feedback is probably being used in some form. (On the other hand, pre-distorting the signal in an inverse complementary fashion can lower distortion without creating a new feedback mechanism. See " Complementry Inverse Distortion Cancellation" for details.) Sometimes, however, it is difficult to see a feedback loop. In the following circuit, is there a global feedback loop or not?

Caution: Lethal voltage at output. Not recommended!

More ð