Transformer utilisation with choke input filter

[Circlotron]

Say we have a nice 100VA transformer powered up and sitting on the bench in front of us. To get 100 watts out of this 100VA transformer without overloading it we would have to have a resistive load, or at least something that draws a sinusoidal current, e.g. a PFC stage. If we put the transformer secondary into a bridge rectifier then straight into a large filter capacitor the current draw will be very peaky, only occurring near the highest point of the AC voltage waveform. The most we will be able to pull out of our transformer without overloading is about 60 watts.

If we then put a large choke between the bridge rectifier and the filter capacitor the effect will be to lower and broaden the current pulses coming out of the transformer secondary, and so we will be able to utilise the transformer more fully. If we make the choke inductance really big the effect will be to make the transformer current almost constant during each half cycle, so the transformer will be delivering a near square wave current despite supplying a sinusoidal voltage.

This being the case, would not the transformer resistive losses be lower than if it were supplying the same wattage into a purely resistive load because the current is spread evenly across the whole half cycle? And consequently, might we be able to pull somewhat more than 100 watts from our 100VA transformer under these conditions?

[T3sl4co1l]

Power factor with choke input can be in the 0.8-0.9 range, IIRC.  Better than cap-input, but still less than full.  This is easy to see because current flows through the transformer even when voltage is low, whereas at PF=1 current is proportional to voltage.

A resonant or semi-resonant filter can do better, but DC regulation is much poorer.  Passive PF tends to be harmonic filtering, producing a compromise between acceptable regulation, and filtering the strongest harmonics (lowest, 3rd and 5th most importantly, etc.).

Tim

[mag_therm]

Yes, an "infinite" choke would be beneficial for single phase line commutated rectifiers, but was not often used because the size of the choke will aproach the size of the transformer. The benefits were not only to the transformer and rectifier, the upstream power quality is improved compared to a capacitor input filter.
The disadvantage of a choke input from an uncontrolled (diode) recifier is the overshoot of DC voltage on sudden reduction of load.

The choke input filter becomes practical at > 3 MegaWatt where pulse numbers of 6, 12, 18, and 24 are used to reduce line harmonics.
Higher pulse number allows for smaller DC chokes.
Related to choke sizing is the the turndown to the discontinupus limit (larger choke is needed for continuous current at minimum load.
This applies to motor control and battery chargers.

Here is a qucs model I made some years ago looking at conversion of an old HP bench supply to choke input filter. Note the square current pulses along with voltage over shoot at end of trace.
https://app.box.com/s/gzm2ui4d9t4e4hcb9ud3jlkhfh1s07j4

[Kleinstein]

Like transformers chokes get better (less loss, improved power to weight ratio) when they get larger. So the choke can work good at high power and less for low power.

The sinusodial current of the resistive load is still the best one can do to get the most power out for a given RMS current. A more rectangular current would draw current at low voltage with little return. So there is not such thing as PF > 1