The circuit design was fairly simple, given the electronic shortage, there was not a ton of options as far as IC chips went. I used a fairly common Lipo charging IC, along with a TI amplifier chip good for 120mW per channel (Way more then I needed, but it led to a simpler design vs op-amps etc.)
I then scaled the resistance down for a typical Baxandall tone control from 500k to ~50k impedance, since the source is the power amplifier output which is very low, and the lowered resistance would reduce any noise. Although given a 1V/V amplification scheme and the large initial voltage, noise is not really a concern here.
Worth noting the THD was also considered, the TI chip quotes 0.02%, but in practice is worse given my low load. An op-amp could do much better, but again as a tube guitar amplifier is the sound source, the small amount of THD added here is insignificant.
I ordered a cheap aluminum box off Amazon that had slots for a circuit board to slide into it. I used 40mm of the 100mm length for the board, and figured the rest would be mounting locations for a power resistor and inductors etc.
Being aluminum, it will act as a heatsink, so no internal fan is needed.
Also worth noting, Ideally this would have been charged via USB, but I lack the tools to make a clean looking cutout for a USB port. So i instead opted to just power it from 9-12V DC. Future mods may include 2 power jacks, one for normal 9-12V, one wired to support the typical -9V pedal power used for most pedal boards (-9V due to center being negative, as opposed to most general devices and power supplies being center positive.
Ran a simulation of approximate component values I was planning on using. The circuit idea here is from here. It is worth noting this site has a lot of good explanation around guitar amps and tube amps. If you want to build your own, it is a nice reference. The first trough is a high impedance designed to mimic the resonant point of the mechanical speaker, and the high frequency roll-off is again mimicking the increasing impedance to higher frequency due to the inductance of the speaker coil and mechanics.
Modern inductors are designed for high frequency power supplies. It is harder to buy an off the shelf inductor designed for low frequency like audio that can handle large currents on the order of a few amps.
From my simulation, I could tell for a 10W load, I needed to handle peaks of 1.5A at least... In reality more was probably needed, but I also figure if my inductor saturates at higher power levels, that might just be something I live with here. Unlike a power supply, saturation here would just lead to more THD and not catastrophic failures.
Here is a good online calculator I found for this, that makes things very easy due to the large amount of unit options, there is less messing around in excel or on paper to balance everything out
There are 2 options for low frequency inductors:
- Historically accurate iron core inductor (Basically a transformer with a single winding since Iron cores are much better for lower frequency inductors.)
- Find a large low permeability torrid that can handle enough windings to meet the inductance needed (25mH in this case is my worst one).
I initially chose to go with option 2 since it seemed easier, and my inductance was low enough it can be achieved without thousands of turns of wire (Or so I thought). I then spent way to long searching Digikey and manufacturer web pages to find something that was not huge, but met the current and reasonable amount of turns requirements. Turns out the 500uH inductor was doable with a modest torrid. The 25mH inductor on the other hand really needed to be silicon steel like a transformer or choke design. There was no readily available ferrite that met my needs.
The 500uH Inductor:
I bought a 30mm torrid and wound #24 magnetic wire around it, I wound something like 60 turns...Read more »