I just learned the hard way how to blow a tube in this amplifier: Turn it ON without R4 or R10.
I wanted to try higher bias at the MOSFET, as I thought mine was low. A few pages back it was recommended to add a 33K resistor in series with the 220K one to raise the bias, so I tried that today. I started by unsoldering one side of the 220K resistor on both channels. I then added a 33K on one side, and thought I should compare voltages with the other side. So I turned the amp on, forgetting the 220K resistor was still unconnected in one channel. The MOSFET went fully ON and blew up the tube heater on this side, within a couple of seconds. By the time I realized what was going on, it was too late. YAY!
So yeah, this amplifier can blow up tubes, be careful.
On a more positive note, I tried the amp with a 20K Alps RK27 and the noise level was significantly lower. It was so low that the noise at the RCA out of the NFB-12 drowned the noise from the SSMH. I thought the NFB-12 had a 10K pot, but it turned out to be a 20K. I really wanted to try the amp with a 10K, as those are easier to get. I'm debating whether to take a chance and just order a 10K one. There didn't seem to be a problem with 20K, but I'm really concerned 10K would be too low.
(yes I did take the pot out of my NFB-12 to try it in the SSMH.)
This seems to confirm my theory. A higher value pot will have the grid "floating" higher relative to ground, and will therefore be more sensitive to noise. I'd really like if anyone with noise problem give it a try and report back. Note that I originally had a 100K pot, so the difference between 100 and 20 may be more significant than 50 to 20. Anyone can try a 10K pot?