Constant Current Load

History

While designing power supply circuits for various products and projects at Parallax, I realized that I lacked a way to really stress-test the supplies, and characterize their actual specifications. Referencing various DIY projects on the web, I threw something together that ended up being a prototype for an MCU-controlled unit.

Building the Prototype

When I say, “threw something together“, I’m not being hyperbolic. As you can see from the prototype, this was just a handful of fairly common parts thrown together on a BS1 protoboard (sans BS1 module).

As shown, the prototype runs from a single 5V power supply, and the most readily available 5V from most of my development boards is the servo headers on most of them. So the prototype is designed to be powered from a servo header running at ~5V.

This unit is made up of the protoboard, an IRL520 MOSFET, a 1Ω @ 5W power resistor, an LM358 op-amp and a 25-turn, 10K trimmer potentiometer (Bourns 3296). The 3-pin servo cable provides power. The red and white wires are the load.

Theory of Operation

This device provides a constant current load for a battery or DC power supply using a MOSFET in its linear region as an active load resistor. This load resistor is controlled by an op-amp, which is configured to maintain a constant current based on a voltage input to its non-inverting input and feedback from a shunt resistor connected between ground and the MOSFET. This feedback voltage goes to the inverting input of the op-amp.

Since the op-amp is essentially trying to keep the input voltages the same, when you set a voltage at the non-inverting input, the output will adjust itself until the voltage at the inverting input matches. By using the 1Ω shunt resistor, the voltage across the resistor equals the current. So, if you set the voltage input to 500 mV, the shunt resistor will have 500 mV across it and you will be drawing 500 mA across the load.

For testing, I connected the unit to a Board of Education for 5V power. The BoE gets its power from a 9V wall-wart. Since I did not have an available current panel meter, I used my Beckman Industrial DM27_XL to measure current, which means inserting it into the circuit path. Power for the board under test comes from my BK PRECISION 1697 Power Supply, which is set for 9V @ 1A (current limited). The DMM allows me to see the actual current being drawn by the load.

My first test subject in the center is the Parallax HomeWork Board, which has an on-board regulator that was upgraded recently to provide up to 500 mA. Since the on-board regulator does not have a heatsink, except the PCB, this test will characterize how much current we can draw before the regulator starts shutting down from over-current / temperature. I am drawing 400 mA before the regulator starts to thermal shutdown.

My second test subject on the right is the Parallax Board of Education, which has an on-board regulator that can provide up to 1A. The BoE does have a small heatsink, but without better cooling consideration, I get to around 700 mA before the regulator starts to shut down from thermal overload. The regulator used (LM2940CT-5.0) is capable of up to 1A current output, however you would need a bigger heatsink, thermal grease and depending on environmental factors, possible even a cooling fan to get to 1A.

So, as you can see, the Constant Current Load allows you to test the actual current capabilities of a regulator or even a battery in a real-world environment. By adjusting the 25-Turn potentiometer, I can dial the amount of current I want to draw..

THIS ARTICLE IS STILL BEING RECOVERED AND RESTORED. PLEASE CHECK BACK LATER! Why is the article being recovered?

Resources

Discuss this project on Savage///Chats