Simple way to improve Topswitch-jx flyback

1 post / 0 new


This post concerns how Topswitch-jx, the world’s best monolithic flyback controller, can be improved by adding an 8V output mini buck converter at its secondary side…this feature is especially useful for topswitch flybacks with vout >20V..
The main advantages is a far better and more certain feedback loop, and big reduction in standby power.

The improvements from adding this mini buck converter are as follows…
1…Permits “true” type 3 feedback circuit, excellent for voltage mode controllers such as topswitch. (“true” type 3 is explained on page 471 of Basso book “Designing control loops for linear & switching power supplies”.
2…Dramatically reduces standby power draw especially when flyback vout is >20V. This is because the current in the feedback opto’s diode is only multipled by 8V to get a reduced power dissipation. (instead of multiplying by 20V or higher).
The mini buck itself would draw its own operating current from its own output, so that will reduce standby power.
3…Means that the Darlington connection of the optocoupler (as in fig31, page 25 of AN-47) is not needed for high vout flybacks. This is a great advantage, as the gain of the Darlington will vary widely due to poor tolerance of BJT’s, and as you know, the last thing that is wanted with voltage mode control loops is variable, poorly toleranced gain values in the feedback loop.
4…The 8V output of the mini-buck can easily be used as a nice, low voltage bias rail from which to power other circuitry, eg, a micro via a 5V regulator, or an indicator LED, etc etc.

Please could Power Integrations update the document called AN-57? Please update it to include the “true” type 3 compensation network given by Dr Christophe Basso in his latest book (“Designing control loops for linear & switching power supplies”). It shows the true supremacy of the Topswitch-jx product that Dr Basso has chosen to devote an entire chapter of his excellent book to feedback compensation for the topswitch-jx.

As you know, in so far as monolithic flyback controllers go, topswitch-jx is miles ahead of the competition. The voltage mode control, allowing 65% duty cycle without needing slope compensation means that we can arrange for a lower secondary diode reverse voltage and thus use a Schottky there, since Ns/Np can be made less when D is high. (it would in theory be possible with current mode controllers but the competition ones don’t have slope compensation included, so their duty would need to be <46%). Also, the Topswitch feature of widely adjustable peak current limit means that we can make our transformer as small and thus cheap as possible. Add to that the decent level of control current in topswitch, -essential to avoid noise issues especially where other switchers are nearby. Of course, with high vout topswitch flybacks, this large level of control current can increase the standby power significantly unless the mini buck is added as per this literature.

However, in order for Topswitch to really take its rightful place on the podium top spot, we need to include the attached circuit involving a secondary side mini buck converter to provide a secondary side 8V rail for the feedback components to operate from. The expense of this circuit is payed for by the fact that topswitch has the aforementioned peak current setting facility, meaning that overall topswitch is the cheapest way, even with the extra mini buck converter at its secondary.

Anyway, as you know, standby regulations are now down to 0.5W now, so the reduction in standby power brought about by the 8V mini buck is very good.

첨부 파일Size
Topswitch flyback.pdf273.57 KB
Topswitch flyback.pdf273.57 KB