CAN YOU USE (Forward converter for high voltage output)

[bowman1710]

I am looking at designing a 200W low voltage to high voltage SMPS (around 350V output). There is a few options I have been focusing on, a full bridge with ZVS topology and a staggered boost and push pull topology. I thought it would be possible with the forward topology but I have read that they are not suitable for high voltage output from the Wurth Electronics site. I cannot see the reason why, if this is the case can anyone explain?

 

[FlapJack]

Define low voltage.

 

[bowman1710]

18V-30V is the input voltage range.

 

[FlapJack]

Push pull is a popular power supply for lower voltages and higher power because it has only one FET dropping voltage at any time, typical use is 12vdc in and say several hundred watts out.

A single FET forward converter should also work for lower voltages like this but i never see anyone use it. Maybe because a forward has half wave out and a push pull has full wave.

But 200W is a reasonable limit for a one FET forward converter.

Both push pull and forward can be used for 350v out with no problems.

My only other comment is, i typically see push pull being used up to 48vdc for 1 or 2kW and like i said i have never seen a forward used at lower voltage levels. It may be because of higher current pulses for the single forward because of half wave out.

 

[BradtheRad]

This Wikipedia article has tables comparing types of smps.

There a few types not yet mentioned (such as the flyback).

https://en.wikipedia.org/wiki/Switched-mode_power_supply

 

[Orson Cart]

The reverse volts on the o/p diode for a forward converter are quite high at your o/p voltage (need 800v min which are slow) - which is why it is not recommended, a push-pull with full diode bridge on the sec is likely the best option for isolated, or a cascaded boost if non isolated (600v u-fast diodes). LLC can be made to work too if you require high freq of operation.

 

[bowman1710]

Thanks for the replys, much appreciated. Flapjack i thought that it might be current related myself. Orson cart, I can see that being a problem. You said 800V min but the document mentioned thar it isnt good with any high voltage outputs. I do want to avoid LLC if possible, i think that a boost to regulate the voltage cascaded with the push-pull for isolation may be the best option. The power supply may have to switch at 1MHz which will obviously affect which may be best suited.

 

[FlapJack]

Bowman1710, 1Mhz sounds a little crazy. Only the best of the best would attempt to make a 200W 350V 1Mhz converter, the technical problems involved are almost insurmountable.

I am assuming you are a schmuck like me and have no access to exotic parts like litz wire. And or also assuming you are only making one of these. That leaves magnet wire and an operating frequency somewhere between 30 and 70 Khz.

I think Orson Cart knows what he is talking about, follow his suggestion for a simple straight forward design. A push pull with one 350 volt secondary winding into a full bridge.

 

[bowman1710]

Bowman1710, 1Mhz sounds a little crazy. Only the best of the best would attempt to make a 200W 350V 1Mhz converter, the technical problems involved are almost insurmountable.

I am assuming you are a schmuck like me and have no access to exotic parts like litz wire. And or also assuming you are only making one of these. That leaves magnet wire and an operating frequency somewhere between 30 and 70 Khz.

I think Orson Cart knows what he is talking about, follow his suggestion for a simple straight forward design. A push pull with one 350 volt secondary winding into a full bridge.

I work with power suplies for my company and this is one if the issues at work. Litz wire is one of the options or a planar transformer due to size and temperature constraints. Yes i am well aware that 1MHz is with this spec is crazy, hence i am looking at 1/3 of 1Mhz (i.e 333.33KHz). Long story short there are only a few options due to external equipment being sensative to a certain range of frequencies. So for isolated 18-30v (input) 350v(output) 200w you would suggested push-pull to full bridge or just a standard forward?

 

[FlapJack]

This is the best choice.

For isolated 18-30v (input) 350v(output) 200w push-pull to full bridge.

 

[Anna Conda]

Yes, if you really need 333kHz for the power converter then you need to have soft switching of the rectifying diodes - else they will cook trying to rectify 350V at this freq, a push-pull operating close to 50/50 duty cycle has inherent soft switching (ZV turn off) and hence no reverse recovery on the o/p diodes (esp with small caps across the fets to stabilise their o/p capacitance) , to achieve regulation you would then need to feed the push-pull with a buck converter (or a boost if you don't mind having a minimum o/p voltage)