|
Recently I posted a bench top power supply, which I've revised one time for increased power capabilities and now I've decided to revisit the circuit, improving some of the other things that bugged me. The first thing I've changed is where the current sensing resistor is placed in the circuit. The last version used low side current sensing, which I usually avoid, because it raises the ground rail above 0V potential and may cause problems with sensitive equipment. This time the current sensing resistor is placed on the high side of the supply (the positive rail) and thus the circuit required a few modifications to make it work. The original circuit also used a Darlington output transistor configuration which causes a higher BE voltage drop, so this time I used a Sziklai pair, to minimize the dropout voltage. In a discussion about the previous version, I've been noted about the high dissipation the output transistors will have at low output voltages and high currents, especially when you use the higher power version I designed. This is a very characteristic trait for linear regulators, but there are ways to alleviate this problem partially. I've added a winding switching mechanism to switch between different winding voltages, when the output doesn't need the full supply voltage. The winding switching works at around 15.4V and switches to the higher voltage winding and works up to output voltages of 30V. In order to stabilize the mechanism around the switching point, a slight hysteresis is added, which disables the mechanism at slightly lower voltage than the first threshold (15.1V). First close the switch and use the 10k potentiometer in the middle to play around with the output voltage to see how the switching mechanism works. Anyway the circuit is not complete, EC doesn't allow me to fit much more stuff. In reality a temperature compensated reference voltage source is required, some stability networks must be employed, and also a proper relay driver must be added.
|
