KOTriode wrote:

Ok, I just read your "Melquides Remote biasing". This technic can not be duplicate in my amp due to DC coupling, the C3G (I dont use 6E5P) driver stage basically adjust the 6C33C current in the output stage. Just for curiosity, I will build a mono block version Melquiades and compare with my version, I believe I have extra chassis for it.

Romy the Cat wrote:

There are no good ready to go transformer for 6C33C. Tangos are garbage.  The 20 years back Tangos were OK for small tubes but Tango is long gone and whatever crap they label by Tango in fact Chinese transformers agglomerate production. Lundahl, with glass core, are very good in my view but they have small core size. At 200mA gap you will have 20Hz but if you build a full bloom 6C33C then you do not want to stop at 200mA. Your voltage is 200V, at 200mA you will have 40W on plate – way to little power, particularly for bass. I do not know the Sowter but generally I would advise to find somebody who would make for you your own custom transformer. Go for 350-400mA gap and for inductance to support 12-15Hz at 20W. There was a guy at my site who ordered one in Europe, I do not know if he like the results – he did not post any feedback. I guess you are American, so you might want o talk with somebody like David Slagle from NY. He is kind of “project guy” and he might be a good source for not cookie-cutter transformers with fast cores. Anyhow, the sound of nay SET is made at 85% by the sound of OPT. If you used any temp OPT to test your 6C33C then you get temp results and temp bass. I do not insist but I think that it is possible that as you get better and deserving OPT for your 6C33C then your view about the 6C33C bass will change.

KOTriode wrote:

Amp at close to clipping level 20.4W, I am surprised that sine wave looked pretty symmetrical. Still has couple chassis left for the Melquiades, the problem is the 400V supplies which required extra power supplies. This interstage amp is so much simpler and easy to build at the expense of the interstage cost.

KOTriode,

the 20.4W is about right. Hoiwever the absolute power is not so important in my view, even though it might be important in case of lower sensitivity that you use. The total power and the amount of distortions is not so indicative in my view to characterize the operation of amplifier. To me it would be more important to see the image of the be trace at let say 22W. I mean I would like to see the first sight of clipping and I would like to see that with your given load the clipping is absolutely symmetrical. If not, and if the loading of the tube that I use is give to me the right sound then I would plays a bit with current/voltage balance to make sure that the right and bottom clips are happen at the very same time and congruous by shape.

Also, you do not specify the frequency of the trace. Did not go over the conversion of your scope time marking into frequency but looking at the sleight asymmetry of the bottom half I think it was relatively low frequency signal. It always useful to know what the frequency and operating current was as it gives a perspective over the inductance of the transformer.

One more thing I learned as I was working on Milq. I tend to put the amp at fill power, get the very first clip by driving the generator lower and lower, make the clip symmetrical and then trace what section of my amplifiers responsible for the clip. An amp might be clipping by driver stage or by output stage. You need to make sure that the clip you get is purely transformer limitation and nothing else.