Using separate secondary windings of a transformer independently

[BobbyK]

Hi,

   I am trying to build a couple of regulated linear power supplies. What I have at hand is one large toroidal transformer with two separate (isolated, not center tapped) secondary winding. The windings are 25V@5 Amp each, and I'm not planning to draw more than 3 amps from each. I am wondering though, are there any issues if say, I am using one supply, and I suddenly switch on the other supply...will the jump in current drawn (particularly the current to fill up the charging caps of the second supply) "ruin my day" or affect the first supply's voltage? If there is such an effect, how do I avoid it? And is it generally a bad idea to use secondary windings on a transformer in this way?


Any guidance and help is appreciated.

[fcb]

You shouldn't notice any significant coupling effects under *normal* conditions (i.e. bridge-caps-regulator).

There will be a slight drop in the 2nd secondary when you draw load from the 1st secondary - mainly due to copper losses in the primary.

What you have to bear in mind is that a transformer can work both ways, and that noise on the 1st secondary is coupled through to the secondary - there will be attenuation due to the iron-core and that will be frequency dependent.  Also, the 'stiffer' the supply and bigger the txfmr the lower the coupling.

Build it and measure it!

[BobbyK]

Thanks. What sort of frequencies are we talking about (coupling between circuits)? Will I detect the effect just putting a function gen output on one secondary, and the scope on the other secondary (without having main on or loads connected)? I don't expect much noise from the supplies, but would simple input filtering (say, series chokes and a couple of caps of varying values in parallel) (after the transformer) for each winding work?

[Avi]

Since you say "regulated linear", I will assume this is a DC supply. This being the case, each should have its own filter capacitor after the bridge rectifier, if you design it right, it should be able to supply sufficient voltage above the regulator input threshold, for the duration of the input sag from the secondary.

[BobbyK]

Yup, Avi. They are DC supplies, and normally I would choose rather big ones. I guess I will have to experiment with the best value for those input filter caps in this case. The problem, I think, is that the larger they are, the more inrush I will get in the already-on supply, and the bigger I make them, the less I will have to actually worry about that inrush lowering the voltage of the already-on supply.

[Simon]

Are you talking about switching in the second winding in parallel with the first if you need more power ? if so just you both in parallel. Make sure they are identical but they should be.

[Avi]

If you really want to limit inrush current, I suppose you could have a series 0.1 ohm resistor (or some other low value) and a relay connected across the capacitor, which would then switch on when it is up to voltage, and short out the resistor.

[vk6zgo]

You are overthinking things!
Many supplies in equipment do just this.

All you need to do is make sure the windings are used within their specifications,use proper regulation,& "Bob's your uncle!"

[IanB]

Since you say "regulated linear", I will assume this is a DC supply. This being the case, each should have its own filter capacitor after the bridge rectifier, if you design it right, it should be able to supply sufficient voltage above the regulator input threshold, for the duration of the input sag from the secondary.

Yup, Avi. They are DC supplies, and normally I would choose rather big ones. I guess I will have to experiment with the best value for those input filter caps in this case. The problem, I think, is that the larger they are, the more inrush I will get in the already-on supply, and the bigger I make them, the less I will have to actually worry about that inrush lowering the voltage of the already-on supply.

This is possibly a bit of a misunderstanding with filter capacitors supplying a linear regulator. The filter capacitors should be just big enough and no bigger. If you make the filter capacitors bigger than needed you will have a worse power factor on the transformer secondaries (lower efficiency) and a bigger problem with inrush current through the rectifier.

To size filter capacitors you should look at the highest planned load and the worst case drop-out voltage on the regulator, combine this with the lowest expected voltage on the transformer primary, then size the filter caps so that the input voltage to the regulator falls just above the drop-out voltage. The regulator will take care of the ripple and any supply fluctuations.

[rbola35618]

If you have a flyback, you just need to be carefully that you always have a load especially if it is a current mode controlled supply. With a current control mode the supply acts as a current source and if you are not closing the feedback on the unloaded supply, the supply acts as a current source. Unloading the supply will cause the output voltage to peak detect and the voltage might go way up and therefore need to make sure the output filter caps can handle the voltage. Hope this helps

[BobbyK]

Thanks everyone. Both supplies connected to the secondaries will be simple regulated DC supplies. I have done inrush protection on Amps in the past, and if I find some of the old circuits around I might just try it here too, though the caps are not that big, and neither is transformer. Overall, seems I should be safe doing this just limiting current a tad and filtering. I'll get started on the supply asap - thanks again everyone.

[vk6zgo]

If you have a flyback, you just need to be carefully that you always have a load especially if it is a current mode controlled supply. With a current control mode the supply acts as a current source and if you are not closing the feedback on the unloaded supply, the supply acts as a current source. Unloading the supply will cause the output voltage to peak detect and the voltage might go way up and therefore need to make sure the output filter caps can handle the voltage. Hope this helps

Lots of SMPS--especially those used by Sony in the 1970s & '80s will happily operate into an open circuit.
The early Philips TV ones,& modern PC supplies  need an external load.

[rbola35618]

The early television worked with open circuit because they were voltage mode control and therefore the flyback transformer acted like a voltage source. However, if you use current mode control, the flyback behaves as a current source and as your load increases the output voltage will increase higher than what you would expect from the turns ratio especially if the supply's loop is not closed on it.

[ElectroIrradiator]

If the DC load current from your output is expected to be 3A, then the transformer probably isn't up for the job. As a rough guide, then the AC RMS load current on the transformer in a standard linear PSU is typically 2 to 3 times the DC load current delivered by the regulator to the load. The reason for this is the non-sinusoidal transformer current, enforced by the rectifier/filter cap combination.

Please disregard if you have already taken this into account.