Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

creating PWM sine wave (variable freq. AC power supply)

Status
Not open for further replies.

petrv

New Member
Hi,

I have a need for low power (30 W is enough) supply AC 115V/60Hz - This may seem trivial for you US guys but I live in a country with 230V/50Hz inside wall sockets. Of course I do already have a step-down transformer but that gives me 115V/50Hz - most of the time it does not matter but now I need a real 60Hz AC. As far as I know the only way around that is to use an inverter but of course the inverters I can buy here all have 230V/50Hz output - good for most people here but for me it is useless. I have considered even ordering an inverter from the US but these things are usually too big and heavy (300W and more) and expensive (it must be true sine wave) - and add to this overseas shipping....

So I have decided to make one, MCU controlled, with display and configurable output frequency (ideally 47-63 Hz). I am using dsPIC33FJ16GS502, I want to drive MOSFETs (using a MOSFET driver) to create a 60 Hz sine wave using PWM (about 50 kHz).

I am now thinking which would be the best topology - I don't want to use high-voltage DC bus as is one common design I saw (for a few reasons including safety) - I want to PWM the input low voltage and connect the output to a step-up transformer. One possibility I see is to use a full H-bridge,
another to use just 2 MOSFETs driving 2 halves of a center tapped transformer. Any suggestions which one to choose ?

I have already made some low-power tests with the MCU driving a small H-bridge followed by a LC filter (2x 2.7 mH + 4.7uF) but the output waveform did not look very nice on oscilloscope, I could see the sine wave but the high-frequency component from the PWM was still quite strongly visible
(but I hope that the transformer will take care of that as well, acting as a low-pass filter).

I have the circuit on a breadboard now so I am trying to improve it as much as possible before I go to designing a PCB.

Thanks,

Petr
 
30W is so low that you could use a sine-wave oscillator driving a 30W audio amplifier driving a stepup transformer.
 
Yeah and regulating the output voltage by the volume control ? Very funny, I had that idea as well but did not consider it seriously.
 
A linear amplifier will waste 25W making heat. A PWM inverter will waste about 5W to 10W making heat.
 
Yes I know - but I plan to use the ADC in the MCU to verify the output voltage (compare it with the sine wave table in ROM) so I can compensate for differences - the output voltage will not be dependent on the load. That would be a little difficult with audio amplifier I guess :)
 
Audio compressor circuits adjust the gain to keep the output level constant. They usually use a Jfet as an adjustable attenuator.
 
How about a small motor generator set, have you investigated that.
Maybe even a gas powered generator, that way you can get your 60Hz and if you lose power, you can keep the lights on.
 
Last edited:
So what do you need to power that absolutely has to have a 60 Hz sine wave input and draws under 30 watts at 115 volts?
 
Interestingly there are such devices. I want to experiment with Insteon home automation modules. Unfortunately they do not work with 50 Hz - they have a timing loop couting from the last zero crossing to start transmit some predefined time before the next zero-crossing. Of course the timing is wrong when you have 50 Hz instead of 60 Hz and unfortunately while it would be possible to modify their firmware to autodetect the frequency they did not bother and hard-coded the timing to 60 Hz.
 
Sounds like your better off finding the circuit in the unit that taps into its power source for the timing reference then just use an accurate clock IC to produce the 60 cycle reference for it from there.

I have seen a number of alarm clock circuits that have a small capacitor and or resistor that taps into one side of the power transformer for the frequency signal that the clock uses. You may get lucky and have a similar system thats easy to tap into and be able to send the 60 cycle signal right to the controler itself.

Its just a thought.
 
The timing is defined inside a microcontroller these devices contain - easy to modify if you have the source code but I don't, it is proprietary. Even if I disconnect the zero-crossing circuit (I'd need to reverse-engineer the hardware first for that) and connect it to a 60 Hz oscillator to fake the zero crossing it won't work as the zero crossing is also used to synchronize the powerline transmitter and receiver.
 
Hi,

I have a need for low power (30 W is enough) supply AC 115V/60Hz - This may seem trivial for you US guys but I live in a country with 230V/50Hz inside wall sockets. Of course I do already have a step-down transformer but that gives me 115V/50Hz - most of the time it does not matter but now I need a real 60Hz AC. As far as I know the only way around that is to use an inverter but of course the inverters I can buy here all have 230V/50Hz output - good for most people here but for me it is useless. I have considered even ordering an inverter from the US but these things are usually too big and heavy (300W and more) and expensive (it must be true sine wave) - and add to this overseas shipping....

So I have decided to make one, MCU controlled, with display and configurable output frequency (ideally 47-63 Hz). I am using dsPIC33FJ16GS502, I want to drive MOSFETs (using a MOSFET driver) to create a 60 Hz sine wave using PWM (about 50 kHz).

I am now thinking which would be the best topology - I don't want to use high-voltage DC bus as is one common design I saw (for a few reasons including safety) - I want to PWM the input low voltage and connect the output to a step-up transformer. One possibility I see is to use a full H-bridge,
another to use just 2 MOSFETs driving 2 halves of a center tapped transformer. Any suggestions which one to choose ?

I have already made some low-power tests with the MCU driving a small H-bridge followed by a LC filter (2x 2.7 mH + 4.7uF) but the output waveform did not look very nice on oscilloscope, I could see the sine wave but the high-frequency component from the PWM was still quite strongly visible
(but I hope that the transformer will take care of that as well, acting as a low-pass filter).

I have the circuit on a breadboard now so I am trying to improve it as much as possible before I go to designing a PCB.

Thanks,

Petr

it is more efficient to drive the filter in DC and use a LF bridge.

Dan
 
I have already made some low-power tests with the MCU driving a small H-bridge followed by a LC filter (2x 2.7 mH + 4.7uF) but the output waveform did not look very nice on oscilloscope, I could see the sine wave but the high-frequency component from the PWM was still quite strongly visible
(but I hope that the transformer will take care of that as well, acting as a low-pass filter).
Petr

Do you mean the ringing of the mosfets when they turn on and off? Put a small (~1000pF) ceramic cap between the drain and source pins to help with that. Either way, I doubt it would make it through a decent transformer.

What frequency is your PWM at?
 
Sorry, I had read that as 50hz.

Can you post o-scope displays of the HF noise? Or just try the caps and see what they do.
 
Class-D audio amplifiers use an LC filter at their output to eliminate their 350kHz PWM frequency. Your 50kHz PWM frequency is too low.
 
I built a 80Khz VFD for an ACIM. There was a little ringing noise on the output, but because it's for a motor, I wasn't worried about it that much. Between the snubber caps and the motor windings, it's not going to affect much.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top