Wide Input Range Push-Pull Converter

[adnan012]

Hi,

Is it possible to achieve wide input range in isolated Push-Pull DC/DC converter.

Input: 10-40V DC
Output: 100V @4A
Switching frequency: 50-100KHz.

I wan to use TL494 type PWM controllers.

Regards!

 

[Easy peasy]

Short answer = YES, long answer is that you get high stresses on the output diodes at 40V

also - better stick to 50kHz if you ain't done this before

10V to 100V is a 10 to one step up ( plus a bit for losses ) so 150V diodes if a bridge output

now you run at 40V in, Vout = 400V peak from the Tx, then LC filtered to get the average given by the pwm

so 600V diodes needed, 6A or 8A each for your 4A out assuming full bridge into LC

1200V diodes needed for centre tap output from Tx

 

[Warpspeed]

The big problem with push pull circuits will be leakage inductance in your transformer that can create voltage spikes on your switching devices on top of the twice dc input voltage the switching devices are subject to.
These spikes can be of sufficiently high voltage to be a problem, which is why push pull has gone out of fashion, except perhaps for ultra low dc input voltages.

These days, half bridge or full bridge switching circuits are used almost universally, because the maximum voltage the switching devices ever see is never higher than the dc input voltage.
Its much easier to specify suitable devices that will operate safely and reliably.

 

[cupoftea]

You can also use vicor modules. Or just a couple of easy parallel boosters, up to 60V say, and then 2 transistor forward to the 100V. Or a Full Bridge from 10 to 100V.....and use one of those controllers which does all the bootstrap hi side drives for you etc. I think linear.com do them, aswell as texas.

 

[dick_freebird]

To me the big attraction of the push-pull is that it only
needs low side switches (albeit higher voltage for a given
VIN). Back when, with no bootstrap high side NMOS gate drivers
or half bridge, full bridge driver chips this was a big deal.

 

[cupoftea]

Yes, ive heard the push-pull today, has its niche application area, but i dont know what that is.

 

[adnan012]

Thanks for reply.

I want to know the closed-loop behavior of the PWM ic. At maximum input voltage and light load , the duty cycle is going to be very small. Power Mosfet requires some pulse width to properly turn on. So how this thing is managed when the duty cycle demand is very low?

Best Regards

 

[Warpspeed]

The PWM in the secondary of your push pull transformer must go through a full wave rectifier, plus a flywheel diode, into a choke and capacitor to turn the pwm into smooth dc.

Closed loop behavior will be highly dependent on the phase shift through your LC smoothing filter more than anything else.

 

[Easy peasy]

fly wheel diode not needed - as all diodes conduct in the dead time - essentially shorting the Tx sec

whether for CT sec with 2 diodes, or 4 diode bridge, output L ensures this happens ...

with current mode control the output L is essentially eliminated from the loop dynamics - making outer loop volt control faster than for pure voltage reg ( no pk curr control ).