• Bassman 10
Another area that I want to get more ear training in is the sound of power supply sag. I bought Kevin O'Connor's Sag Control Kit which allows you to limit the current put out by the power supply. This amp has solid state rectifiers and plenty of room inside the chassis to install this kit. So I'm still playing with that, but the effect of sag seems to be mostly in how the amp feels as you play it. There is some compression with higher levels of sag, but I notice it more in how the initial attack of a note feels.
Another mod was to remove the standby switch and replace it with an on-off-on DPDT half power switch. This is possible because of the way the power transformer is wired, and is nearly identical to the design of a Marshall Super Lead. The center tap is not grounded, and a full wave bridge is used. So by switching between the center tap and one secondary wire, you can use either half the winding or the whole winding. Using the center tap drops the voltage 50%.
Dropping the voltage means the bias has to be reset, so the other pole of the DPDT switch is used to select the bias set resistor. Instead of using discrete bias set resistors, I used two pots so I can more easily set the bias. So setting up the amp entails choosing an amount of sag, and biasing the amp at full B+ voltage. Then switch to 1/2 B+, and set the current drawn at idle to 1/2 the amount used at full power. This is the idea behind power scaling, and as I understand it, adjusts the transfer curve of the tube so you get power tube distortion, but at a lower volume.
What I think so far
I played out with the amp in a band situation. It is really punchy due I think to the 4-10" speakers in an enclosed cabinet, and was able to get a good blackface tone. However there were times when I couldn't hear it--especially in certain parts of its lower register. It just got lost in the band's sound. My thought is that it lacks midrange.There are several resistors that hit me as non-standard Fender values (the plate resistors for the phase inverter and the resistors connecting the bias supply to the plate) that I think may form a filter, so I'm going to play with them. Even with just 2 preamp tube stages, it gets fairly dirty rather quickly so there's no lack of gain. If anything it surprises me at how fast this happens, and want to look into why this is so with fewer gain stages and a 12AT7 phase inverter. Blackface reverb amps seem to stay cleaner a little longer.
The bright switch is a little too piercing at low volumes, so that has room for experimenting.
The bass channel uses a non-typical Fender tone stack, maybe it's a Baxandall, I don't know. Unfortunately this channel sounds really muffled for guitar. I'm thinking that maybe I'll convert this channel to the 5F6 bassman preamp circuit. I only have two tube stages, so I'd use an IRF820 MOSFET for the cathode follower. RG Keen's website is what started this thought, but it might be cool to have a Bassman that covers the tweed and blackface sounds.
DISCLAIMER: Everything I've done here has not hacked the amp. I eventually gave up on this amp, restored it to it's original circuit and traded it in on another amp.
In my opinion, this is not a spectacular amp. It's ultralinear, which usually isn't a good thing in a Fender amp, and when I played it in the store, it woudn't distort and had little character. But it was on sale for a really good price, and I thought I could learn something in trying to make it sound better. Plus the guys at the Ithaca Guitar Works had let me open it up and go through it completely and I had a good idea of what I was getting.I was only able to find the schematic for the non-ultralinear model, so I altered that version to show what I have in my amp. The power supply puts 520 volts on the plates, and is substantially different from the non-UL models. It's supposed to get 70 watts from 2 6L6GC tubes, but it doesn't sound that loud.
Initial cleanup
In checking it out I found that it was biased really cold. The tubes were only idling about 6 watts apiece, when I would have expected more like 17 watts. The amp has a bias balance control, so I had to swap resistors to get the tubes up to where I thought they should be. (The tubes were branded as RCA, but had "made in USSR" clearly printed on them.) That restored a lot of life to it.
The power supply and bias filter caps were all leaky, so I replaced them along with the cathode bypass resistors. That gave it another push in the right direction. I also got rid of the ceramic disk input cap to the phase inverter, and black-faced the phase inverter circuit.
Mods and tweaks
Then I started to experiment. This amp doesn't have a choke, but instead uses a large resistor in the power supply. I switched the screens so they connected to the power supply after this resistor to get rid of the UL, and while there was a difference, it wasn't huge. I'm not adverse to UL operation--one of my homebrews has the option of UL or pentode operation, and I prefer the UL sound. One difference between this amp and my homebrew is that the Bassman 10 uses a lot of negative feedback.
Getting rid of the negative feedback helped a lot. There was a little too much distortion, so I ended up changing the resistor from 820 to 2K2. So at the moment I've got it wired as UL with much less NFB. I'm testing the idea that UL isn't necessarily the villain, it might be UL plus NFB that makes the amp sound constrained. After I get some experience with this sound, I'll wire it as non-UL and try to form an opinion.