A bit of a blast from the past, but I don't think Dave ever managed to complete this teardown due to the charging base being potted. I realised that I had a Lidl charging base for a rechargeable PIR night light / torch in my pile which isn't potted. Although the circuit may not be identical, it shares the same centre spigot arrangement, so it is likely that the coupling coil / inductor is similar.
In the Lidl torch, the coupling receiver is an identically sized inductor, but with a smaller number of turns of thicker gauge wire, as would be expected. I tested it with an old Braun toothbrush, with its larger diameter air cored coil, and the charging performance seems to be the same, so it's likely that it is a similar circuit.
The charging base uses a single transistor oscillator with a capacitive mains dropper. It's likely that the supply voltage to the oscillator varies quite a bit depending on the magnetic loading, the reservoir capacitor is rated at 400V and the transistor has a V
CEO of 450V. It's a bit surprising that such a simple circuit can tolerate such a wide load range without running into problems. Operating frequency is 143kHz off load and 166kHz on load (Braun toothbrush). The waveform seems fairly clean with some flattening at the base of the sine wave.
Anyway, the schematic and teardown photos are attached. I included most of the component values - it didn't seem worth desoldering the MLCCs at I'm thinking of using it as part of a high isolation PSU converter (
https://www.eevblog.com/forum/metrology/power-supply-for-voltage-references/msg3398466/#msg3398466). I did note that the mains caps (and specifically C8) are not X rated, but at least look to be reasonable quality and are down-stream of a low current fuse. C8 also lacks a discharge resistor, relying on to current drain of the oscillator to bleed off its residual charge.
Hopefully of interest to someone.
EDIT: Just looking back again at Dave's video, I noticed that the charging waveform on his Braun base was at significantly lower frequency, around 22kHz and with significant ringing. Maybe the higher operating frequency of the Lidl base accounts for the lack of observed ringing.