Sennheiser, but I have heard and very much liked their HD580 model, which just might be the best choice for the tube enthusiast, as it represents a much easier load impedance (300 ohms); the 32 ohms from the Grado is just brutal for tube based circuits.
  Now for your circuit. When designing a power amplifier, design backwards. (And, yes, driving a 32 ohm load requires a power amplifier.) Start with load first. How much voltage? How much current?
  The Grado headphones are fairly efficient: 1 volt will do for most listeners. Now 1 volt into a 32 ohm load means a peak current of 30 mA will be needed to drive these headphones. In terms of watts, 30 mA is miniscule, just  14.4 mW (W = IČRL / 2), but in terms of the current limits of most small signal tube, such as the 12AX7, it is huge, as for example 15 12AX7s would be needed to deliver that much idle current. Because the 6922 is much more robust than the 12AX7, many less will be needed, 12 less in fact. Three triode sections of a 6922 placed in parallel can continuously conduct a total of 30 mA, without excessively shorting their life. Without feedback, the Zo will be 30 ohms.
   The catch-22 of triodes is that if we increase the current demands on the tube, we must also increase the cathode-to-plate voltage so the greater current flow can be realized. When both the voltage and the current go up in value, so too will the dissipation. A sane limit would be 100 volts across these output triodes.
  The circuit was designed for use with only a single polarity power supply, as a negative supply would be difficult to build without solid-state rectifiers. It also uses a noise canceling technique of not shunting the second grid of the Common Cathode stage to ground through a capacitor; instead, the ratio of power supply noise that makes it to this grid, when amplified at the plate, nulls the noise at the plate. The zener diodes are at the output to protect the headphones from excessive voltage swings at turn-on and turn-off.

                               // editor


Subject: Thanks, thanks, thanks!

   Thank you for writing such a wonderful webzine. This is what the web is all about man, you've hit the nail on the head. Geeks helping each other out, free information for advertisement viewing. With a webzine of this quality, you should have no problem attracting any advertiser you desire.
   Now for my request: I've looked over the archives, both webzine and circuit of the month, and wanted to request a combination of this month's common cathode line stage and the headphone amplifier circuit.
   My desire is for a non-inverting headphone amp, with low output impedance. I want to drive a pair of Grado's. I like the 6922 tube and don't mind using several of them to achieve low output impedance. I would also like to see a purest approach (i.e. no solid state diodes in the circuit, and little or no feedback).
   Thanks a bunch for you time and effort, it is appreciated.


  You're welcome, welcome, welcome.
  You are not alone in wanting to do justice to the Grado headphones; they are wonderful. I own Stax Lambda Pros and drive them with a direct coupled tube amplifier I designed. If you have only heard these headphones with the Stax step-up transformer or even with the Stax hybrid amplifier, you have not heard them. I love them, as no sound reproducer I have heard (whether it be headphone or speaker) has revealed as much detail and nuance as these gems, but I would love to pick up a pair of Grado's best headphones, as the $600 model has the best bass and lower midrange I have ever heard from a headphone. One caveat is that Grado headphones weigh more than others (thus, the best bass), which limited my listening session to only one hour before my neck gave out. I have not heard the latest from

Common Cathode line stage / headphone amplifier, with SE output stage and a power supply noise canceling feature.

pg. 16

< PREVIOUS   Copyright © 1999 GlassWare   All Rights Reserved