Power Supply Circuits

DC Variable Power Supply 1.25V to 22V 0.5A

Often when you are doing experiments you will find that you need a different range of voltages to power different circuits or to test components such as light bulbs. Fortunately you can build a simple variable power supply for DC circuits which is based on the variable voltage regulator, the LM317. Have a look at the circuit below:

You will need a 24V DC 0.5A power pack which can be unregulated since the LM317 will do the regulation for us and using a ready made power unit will keep the circuit much safer than if we had started from the mains side. My particular unregulated supply was as high as 32V without a load so keep in mind that when using this circuit it may be possible to get a higher output voltage (>22V) with low current draw.

The circuit is protected with a 0.5A fuse (quick blow) but of course you can use a lower one if need be and an LED indicates that the unit is on. A downside of using an LED with the 24V power supply is that a high wattage resistor (R1) - 1W - has to be used so there will be energy lost in the heating of the resistor.

The other key components include VR1 which lets you adjust the output voltage from 1.25V (the LM317's internal reference voltage) to 22V (the LM317 needs at least 2V at its input above the output voltage). Keep in mind that for the LM317 to be able to output 0.5A it will need a suitable heat sink and the input power supply must be able to deliver at least 0.5A. Although the LM317 has built-in protection against certain fault conditions, D1 helps prevent harm to the voltage regulator should power continue to be drawn while SW1 is off. D2 helps protect against reverse current at the output.

After building the circuit and even before applying power it is a good idea to check over the circuit and make sure you haven't connected anything incorrectly. In particular, check that IC1 is connected correctly and that the diodes and the electrolytic capacitor are connected with the right polarity, and that there are no for shorts.

A simple way to test that the circuit can deliver is to power a load from it but rather than use, for example, a light or motor, a power resistor can be just as good. If you look at the following table you will see three tests that I did:

To test the power supply I connected a multimeter across the output and set it to volts. For measuring current, I connected the output to 1 of 2 power resistors (see table above) with a multimeter in series set to amps (it's a good idea to use the amps setting rather than mA which on a typical multimeter is usually very low: ~ 300mA). Before taking the measurements I chose a desired voltage and work out the current and power based on the selected load resistor. Once I had recorded the measurements I checked the calculated values against what I had measured and found they were close enough. Knowing what the actual resistance and voltage values were I was able to calculate again and found that using the actual measurements the results were right.

If you do the tests yourself be sure to set the voltage first with the aid of your multimeter before bringing the load resistor into the circuit. If you look at the 4V/10R test you will see that even at a low voltage a relatively high current (0.4A) flows because of the low value resistor. This is why power calculation is important as in this case a 4W resistor was needed and we can judge that the resistor would get quite hot even if suitably rated. Looking at the 1K tests we can see that because of the high resistance a lot less current flows through the load resistor even at a higher voltage (20V) and the power requirement is a lot less.

All content of this and related pages is copyright (c) James S. 2016