The fixed ratio DC/DC converter  ... !
pwrtrnx.com

The fixed ratio DC/DC converter ... !

Another very useful converter approach due to the soft switched nature from no load to full load, able to be used on:

  • Push - pull
  • Half bridge, &
  • Full bridge.

The turns ratio is set to deliver the required output volts, or range of volts, given the input range is fairly narrow, typically +/- 20% maximum.

The converter is run at maximum pwm, less a small amount of dead time for commutation.

For a capacitor on the output (i.e. no or very little output choke) it is advise-able to have a short soft start ( pwm from zero to maximum, say 50 or 100 cycles ) - this limits the peak currents in the mosfets and output diodes - that are otherwise limited only by L-leak & winding R.

The transformer is necessarily gapped ferrite, or a low mu material to increase the magnetising current (i.e. lower L-mag) as this gives a goodly amount of turn off current in the mosfets and allows a smooth and complete transition of voltage from rail to rail on the driving side.

Given that, at no load, you only have the I-mag for soft commutation, and at full load you have the I-mag + the energy from I-load in the leakage inductance of the transformer - you may want to allow differing dead times from no load to full load, longer dead times at no load and shorter at full load.

Or - you may find you can put up with a small amount of low voltage hard switching at very light loads.

Ideally you have just enough stored energy in Lmag at no load such that you can set the dead time to that which just gives a ring to the other rail at turn off - and you will then be assured of soft switching all the way to full load.

Where Imag is quite high or the load quite constant you may want to slow the transitions even further by adding some caps across the mosfets. This gives even lower RFI/EMI and softer, rounder turn off transitions.

This is also seen on the diode side where the slow transition does not hard commutate the output diodes but rather the diode turn off is very soft indeed with no voltage overshoot or ringing.

This is really the huge but un-sung advantage of the fixed ratio converter.

The currents in the transformer tend to be quite square giving very low rms conduction losses - the leakage inductance is now fully utilised in providing rail to rail transitions at higher loads - and is not a source of losses (or annoyance ) anymore.

No snubbers are required - either on the mosfets or the output diodes - which makes the circuit very tidy.

As is often the case, around 100kHz is a sweet-spot operating frequency for this converter, but can be run from 20kHz ( for voltage multiplier outputs ) up to 500kHz due to the soft switching of all the semi's.

IGBT's, and even BJT's can be employed if you wish - especially with larger D-S ( C-E ) commutation caps, but you will suffer increased on drop and possibly turn off storage times.

This approach is possibly the best use of the push-pull converter as all the leakage inductances in this type of transformer are usefully employed in generating a resonant transition, rather than working against the designer.

A proven useful approach for voltage multipliers, pre-regulators, and where an application allows a range of Vout that is proportional to Vin.

Finally - if you use mosfets on the sec side - you have a bidirectional converter which can change power flow seamlessly ( and very quickly ) depending on which side has the higher ( reflected ) voltage - at very high efficiency ...

For more useful ideas and problem solving - pwrtrnx.com

Ive also used this to good effect. Self oscillating version. No snubbers required. Very efficient, low noise.

Daniel Krones

Electronics engineer, experienced in biomed, consumer products and aerospace

4 年

Nice write up Colin! I did a similar design a number of years ago and it worked very well. Conversion efficiency approached 99% at 50% durty cycle. I had one problem due to using dual packaged mosfets. I had some cross conduction on the gates from one half of the bridge. I actually got significant sized flames coming out from the mosfet. Was eventually able to fix that with some diode magic but it was exciting while it lasted...

要查看或添加评论,请登录

Colin J. Tuck ( Snr VP Global Corporate Engineering )的更多文章

社区洞察

其他会员也浏览了