Smart DC Tester Better Than A Dummy Load

Testing DC supplies can be done in many ways, from connecting an actual load like a motor, to using a dummy load in the manner of a big resistor. [Jasper Sikken] is opening up his smart tester for everyone. He is even putting it on Tindie! Normally a supply like a battery or a generator would be given multiple tests with different loads and periodic readings. Believe us, this can be tedious. [Jasper Sikken]’s simulated load takes away the tedium and guesswork by allowing the test parameters to be adjusted and recorded over a serial interface. Of course, this can be automated.

In the video after the break, you can see an adjustment in the constant-current mode from 0mA to 1000mA. His supply, meter, and serial data all track to within one significant digit. If you are testing any kind of power generator, super-capacitor, or potato battery and want a data log, this might be your ticket.

We love testers, from a feature-rich LED tester to a lead (Pb) tester for potable water.

31 thoughts on “Smart DC Tester Better Than A Dummy Load

  1. Looks like a sick device! I wish my electronic loads had logging capabilities. Now lets just upgrade the fets so it can handle a few hundred watts
    Nice to see people starting to switch over to the blue pill style ARM boards as well. That value is hard to beat. I’ve been meaning to go down this road with my own projects for a while now.

    1. It’s not the FETs that limit the power, it’s the heat sink. I agree that a couple hundred watts would be handy, but that’s a real trick to implement. You’re talking a big heat sink with cooling fans, and temperature monitoring to shut down if the fans stop blowing.

      1. I made one with logging and transient response testing capabilities. Granted, i cheated a bit by hooking it but to the Analog Discovery for signal generation and oscilloscope http://www.electrobob.com/psu-burner/ . But it should be easy to scale it up to more power. I agree with what others are saying, switching MOS transistors need to be highly under stressed to last enough. I think that in practice you can be cheaper by using more switching mos transistors rather than buying the linear ones.

      2. One design I saw used a sort of boost converter to dump the power into an array of fan cooled filaments. I’ll see if I can find it again but it could handle several kilowatts if memory serves, but had a more limited (>3V, I think) voltage range.

      1. It is all about how the die is optimised. Switching FETs are optimised for operation in saturation. This can mean that if they are operated in the linear region they may suffer failures due to things like second breakdown or they need to be heavily derated. To m

      2. What Wonbat says!

        Operating a MOSFET in its linear region that was not designed for it may result in thermal runaway:
        https://en.wikipedia.org/wiki/Thermal_runaway

        Kerry D. Wong wrote about it here:
        http://www.kerrywong.com/2016/10/08/linear-mosfet-and-its-use-in-electronic-load/
        (Hackaday posted about his electronic loads here:
        https://hackaday.com/2013/10/28/building-a-dc-constant-currentpower-electric-load/
        and here:
        https://hackaday.com/2017/02/28/beefy-100-amp-electronic-load-uses-two-mosfets/
        )

      3. Thanks to Wombat and LonC for the replies…i will have to look into that further.

        To go a little more down the rabbit hole, I thought MOSFETs had a positive temperature coefficient; meaning that they were immune to thermal runaway…but it appears that the wiki suggests differently. I haven’t ever really used them in the linear mode so it’s all new to me.

          1. I don’t have a specific part for an example, but I’ve always liked IXYS MOSFETs. They are very rugged. Their web site isn’t the best design so I don’t see a way to post a link directly to the linear MOSFETs, but go to http://www.ixys.com and click the Power Devices button, then Discrete MOSFETS, and look for N-Channel: Linear Power MOSFETs with Extended FBSOAs.

          2. Every load I’ve looked into with specs more than a few amps, and longevity find, end up with IXYS linears.
            I’m looking for 24V upwards of 50Amp loads by the way. Heatsinks the size of your granny’s nightstand.

  2. I’d like to build one that can source as well as sink, and simulate a charge/discharge curve for different battery chemistries.

    There’s an uncomfortable point in the development of a battery charging system, where you aren’t sure it works yet but still have to hook it up to an explosion-prone battery in order to find out.

    1. You need a 4 quadrant supply basically. Was thinking of making one. It seems a LT1970 can be a good approach. There is a high power design and they had some notes on how to control via DACs. That is if you can actually buy it from anywhere…

    2. There are several battery chargers, which can also discharge the cells. And they have a serial data port where you can monitor voltage current, capacity E.g. for RC-model batteries at Hobbyking. But often the discharge power is very limited. Mine can charge up to 50W but discharge is limited to 5W.

  3. Using a readily available and CHEAP small outline Arduino clone of some kind as the plug-in controller for a project is smart. I wish more people would do that instead of integrating a microcontroller IC into their design.

    1. Thank you. The circuit does not even have a DAC. It uses PWM to make an analog input for the constant current circuit. But I had to reduce the resolution of the PWM from 16 bits to 10 bits to get is running at high enough frequency (70kHz) to use a simple RC low pass filter

  4. I have lowered the price of the standard electronic load (with 18W heat sink) from 42 to 39 USD to make it cheaper than the Sparkfun Variable Load. And I have added the option for extra lage heat sink (31W) for additional 10 USD. I hope you like it.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.