Using 110 volt appliances with 220 volts     Sometimes someone who lives in a country with 110 volt mains travels to a country with 220 volts and want to use some 110 volt appliance with 220 volts. The obvious solution is a transformer but they are too heavy and bulky, especially for things which consume more than just a few watts. I will explore here some possible solutions.     First I have an adapter sold by Radio Shack. I think it is not a very good solution but it will work for something like a hair dryer. It is rated 1600 W maximum for not continuous use.     Basically it is a circuit similar to a dimmer. A thyristor is switched on part way into the half cycle. The difference is that in this case the thyristor is triggered by the downward slope of the input wave and so the duty cycle is always under 50%. I measured an RMS of about 95 and a peak of about 265 volts.     For a lamp or hairdryer or iron or any other resistive load the RMS is what counts so here the adaptor comes in quite a bit low. But an electronic device often has a bridge rectifier and capacitor which will charge to the peak voltage. This adaptor would probably burn most of these types of devices because the peak voltage of a 110 volt sine wave is about 155 and this goes up to about 265. 100 volts over. Ouch!     Of course, we could modify the circuit a bit so the thyristor fires later and we get a peak of about 160 volts. This would work fine with an electronic device like my printer but the RMS value would be a measly 40 volts and would not be useful for resistive loads. Another possibility would be a circuit controlled by a switch (FET or bipolar) which would conduct when the input voltage is under 160 volts and cut when the voltage goes over 160. This gives us four peaks per cycle instead of two. So much the better. In fact, it is easier to implement after the rectifier bridge: Note that this circuit will automatically adapt to any voltage between 110 VRMS (155 peak) and the upper limits of the switch and other components. You can plug it in to 110 or 220 without switching as it will adapt automatically.     One step more complex is using PWM chopping at high frequency and then filtering. This way we can convert a sine 220 volt signal to 110 volt true sine wave. But if we are charging a capacitor then it is even better to use this PWM chopping to maintain the voltage to which we want to charge the capacitor. This requires a smaller capacitor and improves the power factor. But we still have the problem that if we adjust that wave form for the correct peak value, it will have an RMS valeu which is too high. So I have come with a relatively simple idea using PWM chopping.     By using a step-down chopping circuit with PWM and filtering we can get an output voltage which is a fraction of the input voltage and this fraction is the same as d, the duty cycle of the PWM. If d =60% then Vout = 60% of Vin.     My design would modulate d in the following manner: As long as Vin<190 volts, then d=84% and as Vin grows over 190 volts then d is decreased linearly with the increasing voltage until it reaches 36% when Vin peaks at 311 volts.     In the following graph we can see Vin in red, the duty cycle in black and Vout in blue. Note how Vout Peaks twice per semicycle at 160 V. This is very good for appliances which have a capacitor filter and which charge it up to this peak. They will still charge to the same voltage but now they have twice the number of peaks per cycle.     But the good thing is that the RMS value of this wave is 120 volts which means it will work fine with appliances which depend on the RMS value of the input.     The control function for d is very simple as it is a linear function of the input voltage Vin with a max set at 84%. There is no complex feedback circuit as d is a function only of the input. Vout Peak = 160 volts, same as 120 Vac sine wave.
Vout RMS = 120 volts, same as 120 Vac sine wave.
Now we only need to design the circuit which will do it!

 OK, I have started some design work and here is what I have done so far: a relatively simple step-down chopper controlled by a 555 IC.     The idea is to cut the PCB track between points A and A' and insert the simple circuit shown below which would be mounted on a separate, small, PCB of its own. This assumes the original diodes of the bridge can withstand the 220 V which I have not checked but it would be easy enough to replace them.  With a model simulation this is what happens. The red is the input, 220 Vac on the left and 120 Vac on the right.
The green is the output which is limited to about 150 V in both cases.

I still have to resolve a number of issues. The main one is driving the FET. I have included an ideal transformer which isolates the control circuit from the FET but building such a transformer will be tricky. I think the main problems I would face would be designing and building the transformer and the inductance. Everything else looks quite manageable at this point.
One thing to take into account is the frequency and duty cycle range:

 Vin Ton Toff T F d <130 V 84.7 6.3 91.0 11 93.1% 310 8.1 6.0 14.1 71 57.2%

Toff does not change much and is always about 6 us but Ton ranges from 8 to 85 us and so the duty cycle, d, ranges from 93% down to about 55%. The frequency range is from 11 to 71 Khz.
I think the main consideration when designing the FET gate transformer will be that it should transmit without distortion the long, rectangular pulses, up to 85 uS duration.
A useful thing to add to the circuit would be a crowbar protection with a triac which will short the input (and blow the fuse) if the output goes above, say, 170 volts.

Autor: Alfonso Gonzalez Vespa