Baxandall DC to AC Converter

604

11:57:50

Flip the switch to turn the circuit on.

The four-transformer abomination is actually one big transformer with three windings, one of which has a center tap.

The component values were chosen more or less at random and tuned to produce an approximate 120V RMS 60Hz sine wave because the schematic provided did not give any component information.

I got this circuit from wikipedia (wikipedia.org/wiki/Royer_oscillator). The Royer oscillator has a very similar circuit topology with the only difference being the absence of the inductor (replaced with a wire) and capacitor (replaced with air or other insulation). The Royer oscillator is a relaxation oscillator based on the saturation of the transformer core and produces a square wave that flips when the transformer saturates. This means the frequency is dependent on the saturation current of the transformer. With the addition of an inductor and capacitor, the mode of operation changes entirely and it becomes a Baxandall oscillator. It no longer operates as a relaxation oscillator and instead becomes an LC resonator with positive feedback. This has the advantage of producing an approximate sine wave which makes it more useful as a DC to AC power converter. It also doesn’t require the transformer to saturate for proper operation. In other words, while the circuit topology and schematic are extremely similar to that of the Royer oscillator, the Baxandall oscillator actually operates completely differently.

published 3 years ago

592azy2circuitdude

3 years ago

Nice circuit. Two questions: Is the 4th transformer (1:7 ratio) there to only boost the voltage up, or does it also serve another purpose? And can the center tap transformer be used? I tried to put it in, but the circuit wouldn't oscillate.

jason9

3 years ago

The 1:7 transformer serves no purpose besides serving as the voltage/power output and producing the correct output voltage. The oscillation frequency of the circuit is highly dependent on the inductance of the center-tap coil and capacitance of the capacitor. The mode of oscillation also depends on the ratio of that coil’s inductance to the inductor’s inductance. In this circuit each side of the center-tap coil has an inductance of 1H such that the entire center tap coil from end to end is 2H. If you change the inductance of the inductor you can see the behavior changes. If it’s too high the waveform may oscillate in amplitude somewhat. If it’s too low the wave may be somewhat chaotic. If it’s way too low or replaced with a wire it’ll have flat bits sort of like a square wave.

jason9

3 years ago

Sorry, I’m mistaken. Each side of the center tap is 250mH and the total inductance across the center tap is 1H (the inductance is higher than the sum of the two half-coils because they share the same core).

jason9

3 years ago

If you try to use the EC center tap it may not work correctly because it has an extra inductance in series with the center tap terminal, which in the case of a 1H end-to-end inductance will be 250mH. This makes it so that measuring from the center tap to one end you’ll get 500mH, which is incorrect behavior for a real center tap transformer.

592azy2circuitdude

3 years ago

Swell. Thank you for the detailed information. I experimented with loading the output transformer to see the output current limit. I placed a 170Vp light to the secondary coil and increased the power. I found that the circuit worked great to about 2W. Above 4W the waveform became very distorted. I'm guessing that is due to the transistors saturating but I don't know. What do you think?

jason9

3 years ago

I don’t really know either, but I found that decreasing the resonator impedance helped a lot, not only to improve the shape of the waveform at idle but also to boost the maximum power output. Decreasing the transformer inductance to 1/10th (setting all transformers to 100mH instead of 1H) and increasing the capacitance accordingly to maintain the correct frequency (72.6μF) made the wave almost a perfect sine at idle and also boosted the output capacity to 25W. Above that it also didn’t have a distorted waveform, it just died out. Unfortunately this meant a greater tank current reaching about 1A peak.

jason9

3 years ago

I tried adding diodes from ground to the bases of the transistors so that the bases can’t go below -0.7V and I found that this increased the circuit’s ability to withstand load at the cost of an increasingly distorted waveform as it goes past 25W that looks more like a sawtooth with alternating polarity at very high power draw. This also required that the transistors had a low emitter resistance.

592azy2circuitdude

3 years ago

I agree: Those changes are a big improvement. I suppose the higher current on the low voltage side is because power in equals power out (ideally). Vin × Iin = Vout × Iout. So if the output voltage is boosting up, it would mean greater current draw on the input.

592azy2circuitdude

3 years ago

Now that I think about it, the high current maybe is from the low impedance of the tank circuit at the resonate frequency.

jason9

3 years ago

Yeah, it’s due to the low impedance. If you look at the average supply current, it’s something like 50mA unloaded, even with the high tank circuit current, but it shoots up to 1A with a 20W load (although the output voltage also drops a good bit so the actual output wattage is less than 20W, otherwise we’d have more than 100% conversion efficiency since 1A * 12V is only 12W). Because there’s a constant (and not very large) ratio between tank voltage and output voltage the tank circuit must necessarily have a significant current through it given that we want 120V RMS out and a fairly low impedance take in order to supply enough wattage. In fact, I believe the maximum wattage of the circuit is likely the tank voltage times the tank current at idle.

Issacsutt

3 years ago

Looks like it can source up to about 26mA (output drops to about 145v-pk though) without getting to wacky, so not too bad, that’s a peak wattage of about 3.77watts

add comment in editor