Snfoilhat
Tele-Afflicted
Not sure if this is a topic that'll grab people's attention like making a big amp smaller has. May be interesting or instructive though since it uses a lot of the same ideas just in reverse.
First: remove the EL95 power tubes, their 7-pin tube sockets, and the 8W output transformer. Install 9-pin tube sockets for a pair of EL84s (6BQ5s).
Next: locate the new output transformer and drill new mounting holes in the chassis, figure out OT leads (I decided to stick with a single pass through w/ grommet).
That's all the chassis work and the big component choices made that the new design has to accommodate: a pair of EL84s in push pull, a 6.6k 20W output transformer, and the original PT that was always suited to a larger amp, and so makes this reversal feasible.
Circuit rearrangement:
It would have been possible to make a list of changes that seemed right and then implement them all at once, but I've opted to take it in steps. I left the very large power supply sag resistors and power tube cathode bias resistor from the old design in place, simply slotting in the bigger power tubes and new OT and seeing where that left the amp with respect to voltages, bias current, operating point. That's the left column, below:
Pretty good looking numbers, except for the low plate current making for a plate dissipation of about 50% of max. So I made one change, reducing the cathode bias resistor value from 300R to 200R. More plate current flows, drawing more current through the power supply, dropping more voltage through the huge sag resistors. The new voltages are shown on the right column, above. Still cool -- about 60% of max plate dissipation.
With each change I'm also looking at the estimates from load line calculator, like this:
That brings me up to date. I'm going to play it a bit as it sits to try to make some memory of how it sounds and then tackle those sag resistors. Modern Vox AC15s w/ silicon rectifiers use a pair of 180R resistors. I'll start by picking something in between the present 500R and 180R, and look at all the numbers again, determine the operating point again, listen again, maybe go another step smaller and repeat. Once the power supply and power amp are in a pretty solid place, then move towards the input, tweaking any preamp stuff that wants tweaking. Going back to last year when I finished the first version, I never loved the voicing. Hoping to get that in a better place for V2.
First: remove the EL95 power tubes, their 7-pin tube sockets, and the 8W output transformer. Install 9-pin tube sockets for a pair of EL84s (6BQ5s).
Next: locate the new output transformer and drill new mounting holes in the chassis, figure out OT leads (I decided to stick with a single pass through w/ grommet).
That's all the chassis work and the big component choices made that the new design has to accommodate: a pair of EL84s in push pull, a 6.6k 20W output transformer, and the original PT that was always suited to a larger amp, and so makes this reversal feasible.
Circuit rearrangement:
It would have been possible to make a list of changes that seemed right and then implement them all at once, but I've opted to take it in steps. I left the very large power supply sag resistors and power tube cathode bias resistor from the old design in place, simply slotting in the bigger power tubes and new OT and seeing where that left the amp with respect to voltages, bias current, operating point. That's the left column, below:
Pretty good looking numbers, except for the low plate current making for a plate dissipation of about 50% of max. So I made one change, reducing the cathode bias resistor value from 300R to 200R. More plate current flows, drawing more current through the power supply, dropping more voltage through the huge sag resistors. The new voltages are shown on the right column, above. Still cool -- about 60% of max plate dissipation.
With each change I'm also looking at the estimates from load line calculator, like this:
That brings me up to date. I'm going to play it a bit as it sits to try to make some memory of how it sounds and then tackle those sag resistors. Modern Vox AC15s w/ silicon rectifiers use a pair of 180R resistors. I'll start by picking something in between the present 500R and 180R, and look at all the numbers again, determine the operating point again, listen again, maybe go another step smaller and repeat. Once the power supply and power amp are in a pretty solid place, then move towards the input, tweaking any preamp stuff that wants tweaking. Going back to last year when I finished the first version, I never loved the voicing. Hoping to get that in a better place for V2.
