Hello all
I’m new to this forum and the content is really informative. I’ve tried searching for MOSFET based inverter circuit design but not found anything with relevant links yet. I need to design a 12V DC to 12V AC converter (i.e. an inverter but one which needs no voltage boost). It’s to drive a low voltage extractor fan which is rated at 13W (12V AC @50Hz). I’ve converted all my home lighting circuits to 12V DC to take advantage of solar power I’m gradually installing. It’s all off grid. But the shower extractor fan I’ve bought (without checking it was DC) obviously won’t work on a DC supply. So rather than looking for another fan (this one is quiet and efficient) I’d like to drive it from DC since it should be straightforward with a MOSFET inverter circuit. Any design guidance please? I’m guessing a 1:1 transformer might be needed but wonder if I can manage without a transformer at all? Thanks in advance.

 

A simple H-bridge (such as Toshiba TB67H450) might work, or it might be buzzy.

 

Hi,
probably, the simplest solution is to use 12VDC to 230VAC Pure sine wave inverter and add 230/12V transformer

 

Probably the simplest scheme would be if you already have a 12 volt DC to AC of any voltage transformer type inverter. then a connection to the transformer primary can provide the 12 volts AC. Actually, using the 24 volts across the whole primary could also work because with the much higher frequency the fan impedance will limit the current.
Even then the hard part will be the 50 Hz transformer.
An "H-bridge" circuit can certainly provide a 12 volt square wave quite easily, which might work with the 12 volt AC fan. I have not experimented driving AC motors with a square wave.

 

Why not just use a 12 volt computer type fan ?

Les.

 

If you want to build a circuit, the most simple is a Royer oscillator. Two power transistors, maybe two driver transistors, plus a center-tapped transformer. The transformer is the hard part, because most power transformers have too high a turns ratio.

The solution for that is two transformers. The first one makes a high voltage squarewave at the frequency you want, and the second one steps that output back down to 12 V -ish.

ak

 

I found this on the internet. It doesn't use FETs but it's very simple. You could use any transformer that has a 12V CT winding. Ignore the secondary and take the AC power from across the 12V winding.

 

Why not just use a 12 volt computer type fan ?

There probably is a significant difference between the fan mounting arrangement inside the exhaust enclosure and a typical box fan.

ak

 

The problem with the Royer circuit is that the output frequency is dependent on component tolerances and completely undocumented transformer parameters. An incandescent light bulb won't care, but a fan motor probably will.

Somewhere on the webbergoogle there are schematics of an inverter that has two power MOSFETs in a half-bridge, driven by a CD4047 oscillator. Having an independent oscillator circuit driving an output stage is way more predictable, stable, and adjustable with a single pot.

https://www.google.com/search?clien...uit-using-CD-4047#vhid=og4bsJmyWcR4yM&vssid=l

Turns out - I saved one.

Besides missing power supply decoupling for the IC, the first schematic has a critical error. R2 should *not* be in series between the DC power input and the transformer centertap. The second schematic has the correct arrangement. Power goes directly to the output stage, then through a decoupling and regulating network to power the chip. This prevents switching noise (and there will be a lot of switching noise) from disrupting the oscillator.

ak

 

View attachment 323241

A 12-0-12V output transformer and a 24V (±12V) supply? Doesn't seem quite right to me.
Shouldn't the earth symbol represent 0V not -12V?

 

The schematic in post #7 has wrong battery polarities and wrong battery voltages.
Do it like this:

 

A 12-0-12V output transformer and a 24V (±12V) supply? Doesn't seem quite right to me.
Shouldn't the earth symbol represent 0V not -12V?

That is bad English. It clearly is a single-rail, +12 V and GND circuit.

The fact that a battery is marked + and - rather than + and GND is a common language issue. Is it -12, or +12 V Return, or NEG, or GND, or . . . ?

ak

 

Keep in mind that an asynchronous motor does not perform well when fed with a square voltage. Its moment decreases. And there is a rumble.This is why I mentioned in my post #3 that it is better to use a sine wave generator.

All the circuits shown here generate a square voltage

 

Keep in mind that an asynchronous motor does not perform well when fed with a square voltage.

With tweaking they can do a stepped squarewave, but I see your point.

My first thought was a simple sine circuit such as a phase-shift oscillator, followed by a 20 W audio power amp module from ebay. The problem is that 12 Vac is 34V peak-to-peak, and adding a boost converter to the circuit to make 36-40 V is a lot of gorp.

My next thought was a 20 W power transformer with a secondary rated for 30 V. Driving that backwards with a 3 Vrms (8.5 Vp-p) sine would give you 12 Vrms.

Oscillator >> ebay power amp >> backwards transformer >> fan

By the time you get through all of that, it's probably cheaper to buy the right fan.

ak

 

Brain Fade! It just hit me ... BTL

A bridge-tied, Class D audio power amp module might hit the numbers with a single 12 V supply. If yes, no transformer.

ak

 

Brain Fade! It just hit me ... BTL

You mean, like the TB67H450? Use an 8-pin micro to produce the PWM signal, to give a sinewave output.

 

One other possibility would be a bridged class D audio amplifier driven with a sine wave. As it's output would be a maximum of a bit less than 12 volts peak to peak the RMS output would only be about 8 volts. This could be stepped up using an auto transformer. This could be made by adding a few turns to a 12 volt toroidal transformer and connecting them in series with the 12 volt winding. You would need to insulate the ends of the original primary winding. I think it would be better just to get a12 volt DC fan.

Edit. I have just noticed that AK has beat me to that suggestion.

Les.

 

Here's the spanner in the works.. .
If it is a shower fan, then it's 12V because it has to be Separated Extra Low Voltage, so in a normal installation it has to be supplied by its own transformer because it is in a bathroom. That doesn't change even though the rest of the installation is 12V it still has to be galvanically separated from the earths of the rest of the system, so it still needs a transformer.
Of course, it depends on which country's regulations the TS has to follow, and he didn't state his location.

 

Thanks all for the excellent range of replies! I’m in the UK, and yes it’s a SELV isolated transformer/fan product I’ve got. I suspect as many have suggested that the fan wouldn’t like a square wave supply so I’m looking for a pure sine wave solution. Maybe a small inverter powering the transformer that came with the fan is the simplest answer. I was hoping for something a bit more elegant but I was probably overthinking it!

 

Thanks all for the excellent range of replies! I’m in the UK, and yes it’s a SELV isolated transformer/fan product I’ve got. I suspect as many have suggested that the fan wouldn’t like a square wave supply so I’m looking for a pure sine wave solution. Maybe a small inverter powering the transformer that came with the fan is the simplest answer. I was hoping for something a bit more elegant but I was probably overthinking it!

I‘m surprised that you are using a voltage as low as 12V.
Do you have an inverter that is running, for example, your fridge?
If so, you could step down its output to 12V with an isolating transformer.