Direct-Coupled Tube Headphone Amp

Here’s an idea for a direct-coupled tube (or tube-MOSFET) headphone amplifier. It is a simple variant on a widely used circuit, so I’m sure something similar is already out there. Conceptual schematics for the tube-MOSFET and tube-tube versions are shown in Figs. 1 & 2. A DC level shifter consisting of zener diodes allows the circuit to be direct coupled, with the output at 0 V DC, while also allowing the tubes to operate at high voltages (200 V), where they work best. I can’t claim any creativity on the level-shifter idea, since similar techniques were used to direct-couple tube operational amplifiers back in the 1950s & 1960s. Back then, neon lamps were used instead of the zener diodes. But I haven’t seen this exact circuit for a headphone amplifier before, and it elegantly eliminates the low-frequency phase shift and distortion from an output coupling capacitor or transformer.

Note that a different method of direct-coupling is used in our Minotaur hybrid power amp (, where level-shifted power supplies allow different portions of the circuit to operate at different voltages, but ultimately the concept is the same.

The output device in Figs. 1 and 2 could be a MOSFET, a tube, or an NPN. I use high impedance Sennheiser headphones, so the tube output would probably work fine for me, but I suspect the MOSFET or NPN output would have lower distortion. With a negative high-voltage rail, the output device could even be a 6AS7 (although that is getting a bit impractical for a headphone amp, assuming you want it to be compact).

Figs. 1 and 2 are conceptual schematics, drawn to emphasize the direct coupling, but the output DC offset stability is likely to be a problem. Fig. 3 is a more practical version, with a modified circuit to control DC offset. A DC blocking capacitor in the feedback loop (C28) reduces the gain to unity at DC, allowing the DC offset at the output to be stable to within ±50mV or less, which is more than good enough.

A direct-coupled circuit also requires DC offset protection circuits, in case a tube or transistor fails, or a tube is removed from its socket. Fig. 4 includes a window comparator which opens the output relay if a low-pass filtered version of the output offset voltage exceeds ±0.65 V or so. There is also a turn-on delay for the output relay.

I’m planning to develop a PC board for this design soon. Any suggestions before I do? I assume there are similar circuits out there? How much current is really needed for low impedance headphones? Is there a better output MOSFET than the IRF540? Thoughts about an NPN (or Sziklai) output version? And note, this basic topology could be used to build a direct-coupled hybrid power amplifier as well. I’ll update this website as the project proceeds, adding design updates and measurement results as I have them.


Fig. 1: Conceptual schematic of direct-coupled tube-MOSFET headphone amplifier

Fig. 2: Conceptual schematic of direct-coupled tube headphone amplifier with 6SN7 output stage


Fig. 3: Practical schematic of direct-coupled tube-MOSFET headphone amplifier (right channel shown).




Fig. 4: Turn-on delay and DC offset detection circuitry





Fig. 5: Power supply circuitry. An 18VAC at 1A, 50/60Hz input is assumed, such as from a wall transformer.



Copyright, Trademark, and Disclaimers

© Copyright Tavish Design, LLC 2015. All rights reserved. No part of this document may be reproduced by any means or translated to other languages without prior written consent of Tavish Design, LLC.

Tavish Design® is a registered trademark of Tavish Design, LLC.

The information provided here is subject to change without notice. Tavish Design, LLC is not responsible for errors contained herein or for consequential damages resulting from the use of this material.