Hey there, I want to boost a low voltage to 220v using as small and cheap a design as possible.
The output current should go up to 200ma.
I'm looking for a dead quiet output and a single digit miliohm output impedance between 10hz - 20khz. I could achieve these specs by building a regulator on the high side but I think it may be more cost efficient to attempt this on the low side. Unfortunately I only know how to do this with a linear regulator.
I tried using an online switching regulator calculator to check for the part values and it came up with a 47uh inductor and a 20u capacitor but in spice it drops most of the voltage at 200ma no matter what duty cycle I use so I'm not sure what to make of it.

So I guess my questions are:
How do I determine the parts values I need to hit my voltage requirements?
How can I design a switching regulator to meet my output impedance and noise specifications?
Any advice for me?

 

For first instance i would use a deadicated chip for smps, and not a 555 chip.

Like SG1524, SG3524,TL494

 

I don't think there are any dedicated ICs that achieve my specifications though, right?

 

Hello,

Do you need 200 Volts DC at the output?
The chips @Dodgydave is talking about are for AC output using a transformer.

Bertus

 

I was going to boost using an inductor. I've heard transformer switching design is an arcane art.
Can I not regulate with this method?

 

Please post your .asc file so we can see exactly what is happening.

 

Here you go.

 

I don't think there are any dedicated ICs that achieve my specifications though, right?

Plenty of circuits using the Tl494...

 

That doesn't really say much about the achievable output impedance of the regulator.

 

The output is taken directly from the pulse transformer, no regulators are needed as it uses feedback control to alter the pwm, so there's no heat waste.

 

Yeah but what is the impedance vs frequency plot? I can build a linear regulator with micro ohm impedance throughout the audio frequency range, I'd like to be able to do something similar with the switching regulator, at least single digit miliohms.

 

I don't think there are any dedicated ICs that achieve my specifications though, right?

The MC34063 would probably do it, and there are online calculators to help with the design process.

I'm kindof thinking your application needs a push-pull forward converter.

 

The MC34063 would probably do it

It's a 40v input max IC with internal switching transistors. Are you sure those transistors can handle the 200v voltage swings?

I'm kindof thinking your application needs a push-pull forward converter.

Why so? Seems bigger and more expensive.

 

Can I not regulate with this method?

You can, but not with a 555.

Looking at the numbers, let's assume a impossibly perfect boost circuit with 100% efficiency. 220 V at 200 mA is 44 W, not a trivial design problem for a single stage non-isolated boost converter. To boost 5 V to 220 V is a increase of 44x. So for every amp of output current at 220 V you need 44 A of input current at 5 V, because there must be as much input power as there is output power. For your circuit you need to switch 8.8 A through the inductor. That's a lot. With real world parts it will be more like 11-12 A, and you will need a power MOSFET rated at 20 A and 400 V minimum. And a heatsink.

Because of the enormous voltage spike required for the circuit to make the desired output, that output will be noisy as hell. You can run this at 100 kHz to push the noise outside the audio band, and filter the doors off of it, but it still will be there.

Consider a low frequency switcher, a variation of the Royer oscillator. This is very common in low cost inverters, using a small, standard power transformer driven backwards. The oscillator drives a center-tapped secondary as the "primary", and the normal primary wires now are the high voltage output. Follow this with a traditional bridge rectifier, filter, and linear regulator.

ak

 

Sorry I was mistaken, now that I think about it, I actually only need about 50ma on the output.
I'm looking at your royer circuit and trying to compare the cost to simply using a dedicated transformer.
120v or 230v is too much AC to hit 200v rectified, can that royer circuit adjust the output voltage?

 

The only way to lower the voltage is altering the windings on the transformer, otherwise you're looking at an opto-couper feedback with a Tl431 zener.

 

What is the current capability of a transformer run backwards relative to its quoted current rating?

 

What is the current capability of a transformer run backwards relative to its quoted current rating?

The same as the forward ratings. The transformer core is a power or total energy element, and doesn't care which way the power is moving. If the secondary is rated to deliver 2 A at 12 V, then the core is designed for 24 W continuous. When driving the secondary as the primary you should limit the current to 2 A at 12 V, 1 A at 24 V, etc. 4 A at 6 V might appear to be another option, but this probably will cause the transformer to run hotter than normal because the winding wire gauge is not correct for that much current.

ak

 

What is your actual application that needs single miliohm impeadance and very low noise?
More info often helps to get better answers.