Support QRP CoA Everywhere

Sunday, 6 November 2011

Workbench variable voltage/current limited supply

I just had a some fun with a circuit that was drawing more current than needed and an electrolytic kept popping so after a rummage in the junk box I thought I'd home brew a variable voltage/current limited supply from a salvaged laptop power supply. It worked and was built Manhatten style, anyway I fixed the problem a bridged joint, on the circuit I was looking at. After use I thought I better put this into something more long term as I am sure I will find use for this and I have a nice salvaged instrument case that it will go into. I had not used the more usual transformer and bridge rectifier but a switch mode supply so I have added a common mode choke on the mains input and the DC output from the laptop power supply into the circuit, which is simplicity itself and comes directly from the LM317 data sheet. I use two LM317 to first limit the current and then adjust the voltage.

Looking at the supply with an oscilloscope (must capture a few images and add them) and after a bit of playing with junk box components I added a 3300uF 63V capacitor to the input and a 2200uF 25V capacitor to the output. I might also add a 250K (ish) resistor across both so that when disconnected from mains the input and output capacitors are discharged. I have several 1000uF 100V capacitors I might add later.

The output is via a pair of combined 4mm sockets and binding posts, I will also add a pair of Anderson Power Pole Connections.

An earth connection has been added that utilises the workbench earth and the circuit ground is also connected to. CARE!!! this might not be recommended for any to copy but I KNOW I have a very good mains earth from the workshop supply at the point it enters the workshop and a separate 4mm earth cable is used about the workshop in addition to the 2.5mm ring wiring.

The input 19.5V 4.22A was selected basically as it was available but the important bit for any such supply is that it should be rated at 1.5 times the output and this would allow this supply to provide 3amps. I cannot remember exactly where I heard this but I have not included a 3 amp option and the supply will limit to 1.5 amps, which is also the limit of the LM317 I have used. I could add a pass transistor and obtain a higher current rating.

I will add the circuit but a quick look at the LM317 data sheets (google for LM317 data sheet) will show the options available.

Setting the current limit:
The current limits are set by connecting the ADJ and the OUT of the first LM317 through a 12 way ceramic rotary switch each position of which contains a low value wire wound resistor selected from a recent rally '50p lucky bag', I could have used 1/2W and 1W carbon resistors too. I also purchased a set from JABDOG. The current limits I have set are 25mA, 50mA, 62.5mA, 100 mA, 166mA, 250mA, 460mA (odd resistor value), 500mA, 1.2A and 1.5A the output current limit is directly related to the formula  I = 1.25/R where R is the value of the resistor in ohms. The output is taken into the voltage stage from the ADJ port with the resistor connected through the switch as described connected between ADJ and OUT.

Note that the voltage at this stage is the supply voltage minus the 1.25V drop across the LM317 this LM317 never seems to get hot.

I tested the current settings by feeding this output into a 12V halogen bulb that will run at 50W so is rated well above the power supply with a current meter in circuit.

Voltage Setting:
The variable voltage is set by the second LM317 by adjusting the resistance R1 that is connected from the ADJ port of the LM17 to the ground so that Vout = 1.25 * (R2/R1) where R2 is a fixed 270 ohm 3W wire wound resistor connected between ADJ and the output port of the LM317. The voltage can be varied between 1.2V and 17.3V.

I allowed the supply to feed the 12V halogen bulb at 500mA and left it running for about 4 hours and then remeasured both the voltage and current levels and saw no major changes. At 12V 500mA the LM317 will need to be able to dissipate the heat so I used TO220 version and bolted them to the front panel. The lower the output voltage the more heat is dissipated.

I have not added a meter to measure the current output as the limit is set by switch, the output voltage is another matter and I have employed a 3" chinese analogue meter that looks a bit retro (I might remake the scale).


No comments: