[2004-10-6]

6sn7 "production" model

6sn7/12sn7 prototype

My buddy Bob Houlton has been helping me develop and improve my guitar amps. He is a fantastic guitarist, and has a great ear for tone. Our current goal is a low power amp that does that "Blues" tone. Of course the Fender Champ circuits do a pretty good job, and sound great when played clean, but there is something about push-pull that just sounds right. But how do you add a second tube when 1 is too much already? The quest is on for low power tubes that sound good in guitar circuits.

Several people have tried using the 6sn7 as an output tube. I recently scored a bunch of really nice 12sn7 and 12sl7 tubes, and so I thought I would give it a shot. I built up a prototype chassis with a 5y3 rectifier and 3 octal sockets. I figured it would be cool to use the 12sl7s in the preamp, and 12sn7 as the output push-pull pair.

Why 12sn7 and not 6sn7? Well, mostly because 6sn7s are getting pricey because they are so common in hi-fi amps. 12sn7s are somewhat cheaper and more plentiful.

My first version used both halves of a 12sl7 as gain stages and had the 12sn7 as a self-split output pair. The simplicity was appealing, and the amount of gain was good, but the self-split configuration didn't sound very good. It got muddy and over-distorted very easily.

Next I inserted a phase inverter in front of the 12sn7. The phase inverter was the classis 12ax7 circuit with a 12sl7 in it. This sounded much better, but had WAY too much gain. Even substituting a 12sn7 for the phase inverter still had WAY too much. But the sound was getting good, so I kept working on it.

I had been thinking about Marshalls and why they sound so good, and decided that 2 of the ingredients in the Marshall tone are the cathode follower that feeds the tone stack, and the long tail pair phase inverter. Since I had so much gain, I converted the first gain stage to a cathode follower, and added Steven Delft's Moonlight tone stack (since this amp is so Moonlight-like). With a 12sl7 doing the phase splitting, this had just enough gain to get nasty. The cathode follower really brought out the brightness. I am also running the 12sn7 at fairly high plate voltage, which tends to brighten things up a bit.

Bob and I were really digging the tone at this point, and the last mod was to add a master volume control between the phase inverter and the output pair. I removed the grid resistors on the output tubes, and replaced them with a dual 500K log potentiometer. Bob really likes having mutiple volume controls on an amp, because it lets you dial in how much distortion you want from each gain stage, and how much distortion you want from the output tubes. After all the most important "tone control" on an amp is the volume knob!

Bob turned up the 12sx7 as a 12sn7 equivalent that gets rave reviews from hi fi builders. We tried CBS 12sn7, Sylvania JAN 12sn7, and RCA JAN 12sx7 tubes as the output pair. The CBS 12sn7 had great bass and the most peircing highs. There was a bit of a split on whether the highs were too harsh, or if they were cutting in a good way. The Sylvania JAN was limited on both ends of the frequency range. Less high and less low end. The low end on this amp is suprizingly good, so it is a shame to lose the bass. The RCA JAN 12sx7 brough back the bass, but the highs were still duller than the CBS 12sn7. If you like the tonal balance darker, this is probablybetter, but the piercing highs of the CBS will have fans.

The 12sl7 tubes have been tan base Sylvania JAN tubes the whole time, and they sound just plain great.