Hi Jay,
I've just started learning about tube amps, so please take everything I say with a grain of salt. I'm also starting a Tramp build, though.
Regarding the arcing issue, I'm tempted to simply leave out the SPDT switch on the transformer primary. As far as I understand, it's the ratio of turns in the primary and secondary that matters, not the absolute number of turns. So I imagine imagine switching an 8 Ohm speaker to the 16 Ohm secondary tap will have the same effect as leaving it connected to the 8 Ohm output tap and switching the tube output to the 2.5K primary tap.
In other words I think (maybe someone with more experience will correct me) we can delete the sparky SPDT switch and happily run an 8, 4, or 2 Ohm speaker in the Tramp using an EL34 by selecting the 16, 8, or 4 Ohm output impedance tap respectively. Of course, this would require an ugly sticker on the chassis panel explaining how to set the impedance switch when running an EL34.
Regarding the cathode bias resistor value, I don't have any insights into how the different tubes will respond to the different resistor values. OTOH, I wonder if we could have all options at once... bear with me through the speculation below:
- In stock configuration, with S4 closed, the bias resistance looks to be 560 Ohms paralleled with 750 Ohms, which comes out to about 320 Ohms (note that 330 Ohms is a standard power resistor value).
- In modified configuration, with S4 closed, the bias resistance looks to be 1000 Ohms paralleled with 750 Ohms, which comes out to about 430 Ohms.
- In either configuration, with S4 open, the bias resistance is 750 Ohms.
By coincidence 330 and 430 sum to about 750. Seems like we could have all three bias options if we replaced the bias resistors with 330 Ohms and 430 Ohms, and wired them in series. Then we could swap S4 for an "on-off-on" type switch, connect its middle terminal to the junction between the 330 and 430 Ohm resistors, and connect its other two terminals at either end of the series combination. When S4 was in the center (off) position, we'd have the full 760 Ohm bias resistance. When switched up or down, it would effectively jumper out either the 330 Ohm resistor or the 430 Ohm resistor, leaving the bias at 430 Ohms or 330 Ohms. I'll try to attach a picture illustrating what I'm imagining.
Attachment:
File comment: Image illustrating the bias idea
schematic.png [ 23.38 KiB | Viewed 9162 times ]
It feels a little irresponsible speculating about this, given how green I am, and considering that I haven't tested anything. Please don't take my word for anything. Perhaps someone with more expertise will chime in...
Thanks.