I feel like I have to apologize for being all over the place with this project.
Scratch-building tube amps is not a good hobby for people with short attention spans, but I am doing it anyway.
It's basically working and sounding pretty good.
What did I do to fix it? I don't really know.
The variety of problems must have been mostly caused by the bad sockets. I also had a 6п14п on the left channel that was sensitive to a tap or flick. I'm not sure if it is the tube or just the socket issues. Remember, these are all
very used Soviet military surplus 6п14п's (just over a buck a piece, including shipping from Ukraine), so I can add that to my list of excuses.
It's not that I am not smart enough about building tube amps, but I guess I am not smart enough to buy stuff that will help me avoid these problems. Tube (and socket) bottom feeding does sometimes result in aches and pains.
I've closed it up, and I am basically considering the project complete.
It looks cool, the LED VU meter works and looks very nice -- It has plenty of power, and sounds at least "OK".
Despite the way the SMPS works with regard to voltage regulation, it does work to run a basic tube amp. It actually emulates a Linear PS in that the output (each output) is dependent on load. There's an adjustment pot, but since it effects all outputs, you basically take what you get.
The SMPS definitely has some advantages. In addition to the wide input voltage, it is also very compact. I think it would be very difficult to build an EL84/6BQ5/6п14п P-P amp in this small of a chassis with a conventional power transformer.
It also runs very cool. With the SMPS frame (Heatsink) bolted directly to the aluminum chassis, there's no way it can get hot. The SMPS transformer does get warm to moderately hot, but I ran it for about an hour today with everything closed up, and nothing reached a temperature that would cause me any concern. In fact it's the coolest running tube amp I've ever seen. Nothing but the power tubes themselves get hot.
This cool operation is also shown in very low power consumption for an P-P tube amp. Here's a screen shot from my power consumption meter (music playing very softly -- idle for all practical purposes):
Another observation about this SMPS. It shuts down upon applying high voltage (after the 30 sec delay) if the power tubes are hot. It seems that it does not like the sudden load on the +300V line if the tubes conduct too quickly. In a way, this is good as it is not a good thing to do to tubes anyway, but what concerns me is that this may mean I am close to the max load even during normal startup/operation. It has never done this during a cold start, and never tripped during normal operation. I will also add that all of the SMPS amps I have built have difficulty starting up when the tubes are hot.
It's not the ideal tube amp SMPS solution I'd like to see, but for a simple power amp that is
at most what is described here, it does work.
Here's a finished picture for now, and I may post better pictures, and some square wave response pictures if I get a chance this evening.
A split load triode stage. Idling with its grid at 25% of available B+. Raise its grid a volt. Cathode follows it( goes up a volt ), and the plate comes down a volt. For a 1V input the plate-cathode change is 2. Now with approximated cathode follower voltage output, we are looking at .8 to .9 volts response to the 1V increase in grid voltage. Sooooo, grid to plate/cathode voltage is usually about 1:1.8-9.
-- Is that I used automotive connectors (Insulated spade lugs) for the high voltage leads. I did this so that the entire top can be easily removed. (Six spade lugs in total -- Ground, 300V, 270V, times 2 channels each).