A Constant Resistance Dummy Load Design

constant-resistance-dummy-load

This constant resistance dummy load has not yet been tested in the real world. [YS] was inspired to come up with the circuit after reading Wednesday’s Re:load dummy load post. That was a constant current load, not a constant resistance load. [YS] started with the schematic for the Re:load and made his changes to arrive at this.

For him the exercise was just to alter the design to achieve constant resistance. He didn’t actually build and test the hardware because he doesn’t really have a need for it. This image was exported from Proteus, which includes a ProSPICE circuit emulator. His slides run through test voltages from 5V to 50V, maintaining a constant 10 Ohm resistance.

When studying this project we needed a little refresher on the different varieties of dummy loads. We found this post very informative about the differences and uses of Constant Current, Constant Power, and Constant Resistance (Impedance) loads.

21 thoughts on “A Constant Resistance Dummy Load Design

  1. Just a little question from someone not so familiar with electronics. Wouldn’t a 10 Ohm resistor give a constant resistance across all voltages? Or is this adjustable? But wouldn’t a potentiometer be adjustable and give a constant resistance?

    1. This is adjustable – or at least, will be if you replace one of the resistors on the left with a potentiometer. The reason to do this instead of simply using a potentiometer is that potentiometers can’t handle much current at all – fractions of a watt – while this can handle many watts, with appropriate heatsinking.

  2. Very neat solution! One note, though – the LM358 has relatively high offset voltage and isn’t rail to rail on either input or output, so that’ll cause some issues driving this circuit at extremes of supply voltage or current.

    1. Rail2rail capability isn’t critical here, because MOSFET’s gate has a threshold voltage of about 1.5 V, as mentioned in datasheet, and maximum gate voltage is around 10 V. So in the voltage range above 10 V we are just in safe operating conditions. But of course using rail2rail opamp can improve operation in “below 10 V” area.

      1. Yup. This is something I really struggled with trying to make the Re:load work at low voltages. Also, lower offset voltage means you can use a smaller shunt resistor, which means less power dissipation through the resistor.

          1. Yup. But I figured that people probably use this as a tool relatively infrequently. Finding that every time you grab it the battery is dead – or having to find a DC plugpack and a place to plug it in – would be a major pain, and, I think, make it a lot less useful.

    1. I’ve been pondering that myself. I think it would suffice to use an opamp configured as an inverting amplifier hooked up to the input voltage, with its output driving a voltage divider (with a potentiometer) for the existing current feedback loop.

      Combine all that with a three-way switch, and perhaps you could have a dummy load that can be configured as constant resistance, constant current, or constant power. Hmm. Version 2?

      1. you’ll need something to do multiplication to get constant power, the exponential of the diode equation can get you some of the way but it will take a bit more to get it working and stable with temperature etc.

  3. Please keep in mind that real circuits like this REALY like to oscillate. It takes quite some patiance or/and math when stable and maintain fast response to like in this case respond to voltage changes.

    1. Agreed. Don’t count on SPICE op-amp models to perform exactly as their real-world equivalents. Still, it’s good practice to bang up circuits in SPICE as personal exercises, even if you have no real need for them. I have dozens of them, and I think it was time well spent.

      To [YS]: Here’s an extension of this exercise I recommend. Simulate this with a squarewave input going, for example, from 5V to 10V. Watch the voltage on the op-amp’s VCC pin, as well as the transient response of the circuit as a whole, particularly when the input steps down. Then add a 0.1uF bypass cap between the op-amp’s VCC and GND pins, and see what happens differently. I haven’t used your version of SPICE, but mine has virtual oscilloscopes in addition to the virtual multimeters; use the scopes instead so you can see transients in better detail.

      1. Loop stability 101: There’s a feedback loop going from the output of the opamp to the gate of the FET and from the source of that FET back to the negative input. If you really want to get technical, you could break the loop and run an AC sim to figure out the gain and phase margin.

        Or, to cut to the chase, put a capacitor on the output of the opamp and connect the other side to ground. This will make that low impedance net the long pole in the loop and stabilize the feedback. You’ll need to play around with different cap values to get to one that’s stable. The larger the cap, the lower the frequency will be on that node and the closer you will get to a stable feedback loop.

        1. most opamps really don’t like to drive a purely capacitive load, you should put at least a few ohm in series with the output

          and while we are at it, when choosing the fet, make sure to check the safe operation area

      2. Yes, the stability here can be an issue. But, as mentioned above, I have no real need in a such device, so this design was only exercise for me. This is why I didn’t pay attention to some details required for stable real-life operation and stopped just at SPICE simulation. :)

  4. Kudos to YS for posting his design, even if untested. Something I have a particular dislike of is spending my time re-inventing the wheel. Even if it is untested, and it might have oscillation issues to deal with, I think that this is a much better starting point for me than a blank sheet of paper. Then it’s enhanced by Rasto dropping the idea of sinking constant power. While I don’t need that for my work, it is a clever idea I’ll not soon forget. Thanks to all!

Leave a Reply to MathiasCancel 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.