I've dropped the "Darling Killer" name, and now call it the "DK-0.1" (DK-zero-point-one).
Here we have a little 9-pin tube the 6350. It is a high pervariance dual triode. When the plates are combined, the dissapation is 7.7W. With a nice set of characteristic curves, this tube seems a natural.

I've tried several configurations both for the output and the driver. Originally designed to use the 6N1P as the input stage, I sacrificed some gain for the sonic qualities of the 6CG7; a tube worth hearing. The average CD in my player drives this amp into clipping with the 6CG7's, so I'm happy.

For the original output design I tried various transformers and configurations. The "final" prototype used the Hammond 125ESE. These were connected with the secondary winding in the cathode circuit and the output was taken from the 10K/8R tap. While this configuration sounded excellent (best I've heard from these trannys) it left the bias voltage exposed at the speaker terminals (common self bias resistor). Of course, I could have wired each cathode to it's own bias resistor and then to the secondary winding ... and I did this for a while, but it sounded better the other way.

I will be releasing the original 125ESE based schematic as soon as I get time, as I think that was a great sounding combo as long as care is taken concerning the voltage on the speaker terminals. I will release this amp's schematic soon after. In addition, I have the design for using the 6N1P if anyone is interested. I also have the bias setup resistors for the input stage and the output self-bias power resistor available if anyone is interested. I mention this because the values are 750k 1/4W, 255K 1/4W and 1250 ohms at 8W and need to be fairly exact. You get all 5 resistors for $6.30 includes first class postage to USA. These resistors will work for the 125SE version and require 360V B+ for the output section and a drop to 300V for the driver section. Write me for details.

Oh Yeah! You want to know how it sounds!
Like I can tell you in print, wanna borrow my ears? LOL
It sounds great and better than the Darling (IMHO) with lower distortion overall and excellent depth and imaging. In a word: Wonderful! Power output you ask? About 1W. A tad less. It clips like a Darling in the early stages, but when driven hard (past initial clipping) distorts more than the Darling does.

Update: 5-4-05
The unit has been completed and pictures are now posted.
I've been calling it the "Darling Killer" because I thought the increased linearity and higher power would blow the Darling away. Not the case. This amps stands in its own right with excellent potential. It has the "Darling" like sound but has better more defined bass. When over-driven it clips faster than the Darling. Grid current comes on faster, and grid stoppers are required for this design.

I created a low impedance drive into the final output. The first section of the 6CG7 develops all the driver side voltage gain and is DC coupled with AC bypassing to the second section operating as a cathode follower direct coupled into the 6350 grids. This probably accounts for the excellent bass characteristics.

The little amp sounds great and makes 900mW per channel - a bit more than the Darling.
Definitely proved itself as worth moving on with this configuration. I will build more of these. Meanwhile, the prototype is for sale. The plus/minus 3db points are 15Hz and 15KHz. Output noise contains zero hum and total noise is at my measurement limits. Just a line of hash on the scope when set at 5mV/div of about 2mV. Excellent for high efficiency single driver systems.

The experiment will continue. I'm going to test a common cathode triode active loaded stage version. Probably will give the 5965 a try here. Using this method, I can incorporate hash noise reduction techniques, although I don't think I will have any luck measuring it!

Stay tuned!
jef

Lots of shadows make the wiring look more complicated than it is.