I’ve always wanted to know what the “tube magic” was all about. There is much opinion in the science of music production, probably because music and its perception is highly personal and subjective. Ive always imagined that since transistor amplifiers were “perfect” with their large amounts of negative feedback, great linearity, and low THD that tube amplifiers must add something to sound that generates their appeal. From the reading I’ve done it has to do with harmonics.

After getting back into headphones because of my kids, I decided to try building a tube amplifier that would drive my nighthawks. More specifically I wanted to build an amp that would be sort of an old school / new school hybrid, one that used circuit boards and SMT components for good EMC considerations, but had no semiconductors at all and no negative feedback.

I owe much credit to Matt at cascadetubes.com, whose design of the 6CY7 amp is what I used with some modifications. This amp appealed to me not only because of the SET topology featuring no negative feedback (second stage does have degeneration), but also because the first and second amp stages for each channel were in their own tube. This should guarantee high channel isolation, and with the output tranformers I can send a differential signal over twisted pair to each side of my headphones since they have seperate plugs for left and right. It basically lets me build a power supply and two mono block amplifiers on top of a metal ground plane, with left and right channel isolation all the way to the headphones.

The most significant change I made to Matt’s design was to add a circuit to float the cathode heater virtual ground anywhere from 0 to 30VDC. In testing I found the most amount of power line feedthrough occured at around 7-15 volts. I set it in the low 20s and the output is completely silent with no input, even at full gain (it was pretty quiet at 0 too though, but I wanted to test it).
I left most of the circuit values alone, and substituted a lot of the packages for SMT or “audio” parts like poly film coupling caps, polymer electrolytics for bypass, pro audio taper pot, etc. I used SMT parts at the amplifier tubes to keep the current loops small, and carefully controlled the ground returns. Heater wires are twisted pair routed under the ground plane. I used Panasonic electrolytics in the power section based on their capacitance, voltage, and ripple current. The magnetics are all what Matt spec’d, with the exception of the Edcor transformers with more low end and 16 ohm output impedance. I also added a jumper that lets me defeat the second stage cathode bypass, I wanted to see what the difference sounded like. It’s supposed to only affect low frequency roll off but affects overall gain and does perform degeneration so has to make a difference. Unfortunately it changes the gain quite a bit so it will be hard to tell dynamically.


Leave a Comment

Your email address will not be published. Required fields are marked *

= 4 + 4

(Spamcheck Enabled)

Scroll to Top