I like the idea of single ended single or double parallel output tubes selected by the flick of a switch. I have my eye on the "littlerock: power amp section here: http://home.polstra.com/amps/littlerock/ The pertinent info bold in blue: "The output stage is a cathode-biased parallel single-ended stage using two 6AK6 power pentodes. The 6AK6 is a 7-pin miniature version of the 6G6 octal tube, with essentially identical specifications. Its plate dissipation rating is 2.75W, and in a typical class A stage (the 6AK6) it can deliver 1.1W into a 10K load with 10% distortion. By running two tubes in parallel, I was able to use a fine Heyboer 5K output transformer that I bought from the nice folks at Doberman Music Products. The power tubes run with 180V on the plates and screens, and 15mA plate current -- values I took directly from the data sheet. I used separate cathode resistors and bypass capacitors for the two power tubes, for two reasons. First, with separate cathode resistors the tubes don't have to be matched as well. Each tube biases itself without affecting the bias of the other tube. When a shared cathode resistor is used, any mismatch between the tubes is magnified. If one tube is "hotter" than the other, then its extra current raises the cathode voltage and causes the other tube to run even colder than it would run on its own. Using separate cathode resistors, I found I could use any two random 6AK6 tubes without any problems. Second, the separate cathode resistors allow me to remove V4 and run the amp on just a single power tube when I want to. This has some interesting effects. It reduces the loudness of the amp, but not as much as you might expect. It also changes the harmonic structure of the distortion, giving much more 2nd harmonic content and much less 3rd harmonic. There is a graph in the RCA data sheet for the 6AK6 showing this. I like the single-tube option enough that I plan to make it switchable. That's easy to add. Just insert a 47K resistor in series with R16, with a switch that shorts out the new resistor when it is turned on. C10 should bypass both resistors. When the switch is closed, the circuit is identical to the schematic as drawn. When the switch is open, V4 is biased into cutoff and is effectively removed from the circuit." So, it appears that he runs the single tube also on 5K OT windings. Without knowing why, I assume that this influences the second harmonic content he refers to. But it also seems an impedance mismatch from the 6AK6 data sheet. So what are the drawbacks running through less impedance? AM I correct that this draws more current from the tube? And does this create tangible risk for tube failure? I am considering using a line transformer as OT which has multiple inputs. I could fashion a switchable impedance at the OT, 8K for a single tube, 4K for two, but then I expect that to negate the second harmonic aspect. Is this correct?