It's a Learning Process - Dummy Load with Display

While I was waiting for the prototypes of the next dummy load kit to arrive, I started imagining what it might look like to have the dummy load calculate and display the power being transmitted into it. I went into my magic bag of random electronics components and threw something together on the breadboard.

Dummy Load with Power Display Prototype 1

This big pile of shattered dreams was my breadboard prototype experiment.

I started off with a Raspberry Pi Pico to use as the microcontroller to do the calculations, a small monochrome OLED display to show the results, and a voltage divider (resistors) to bring the voltage down to something that the Pico could read. It worked, but it was very unstable, and it was only reliable when the transmitter was operating in a specific frequency range. When I went out of that range, all sorts of fun things happened. Fortunately, I'd experienced the joys of rogue RF in the shack before, so I knew what the proble was. Figuring out how to fix it was another matter entirely. I tried using capacitors to clean up the data wires, and experimented with different resistors, but it was a lost cause. If you'd like to see one of the early tests, check it out here: Dummy Load with Power Display Test.

Breadboards are not the best for RF work, so I needed to test this another way. I decided to build a small PCB to hold the components and see if that would help. Since a PCB gives me complete control over the layout of the traces connecting the components, I was sure that it would solve the problem. However, it takes time to design and print a PCB, so I decided to bundle a few experiments into one. The result was much prettier, and much more stable. It didn't work exactly as expected, but I was able to learn a lot from the design, so it was a very uccessful failure.

Dummy Load with Power Display Prototype 2

There are a few things here worth pointing out, so I've numbered them for reference.

1. Input capacitor - This capacitor bridges the two components of the coax feedline. It flattens the SWR curve as the RF frequency increases through the 2m band. Beyond 2m, it increases sharply, effectively putting 2m as the upper limit for this dummy load. With the capacitor removed, it functions exactly as the original dummy load. Below it is a SMA connector that I was testing as an alternative input connector.
2. OLED display - It took me far too long to figure out why this little guy wasn't working. I had the data and clock lines reversed in the PCB. That's why it has a those fancy jumper wires sticking out from the back.
3. Voltage divider - This is the same as the one in the first prototype, just without all the extra experimental junk. It's a simple circuit that brings the voltage down to a level that the Pico can read.
4. Raspberry Pi Pico - By soldering the Pico directly to the board, I was able to keep the traces short and route the traces in such a way as to be shielded from the RF coming into the dummy load. This was the key to making the display work reliably and keep the Pico from randomly rebooting.

What I Learned

Impedance varies with frequency - I knew this, but prior to building this version of the prototype, I hadn't considered how important the variations would be in terms of computing power. If you've read my Measuring Power document, you know what I'm referring to. These are the experiments that led to me creating that document.

At HF frequencies, the measurements were close to what I was expecting, and were reliably so across radios. However, they were very far off once we got into the VHF range. That put me face-to-face with a huge design flaw - without knowing the frequency of the signal, the power measurements were meaningless. I investigated ways of building in a frequency counter, but they were all too complex for the simple project I was trying to build, so I abandoned the idea. If you know of a way that I can do it with a Pico, let me know! In any case, I've decided on a simple and inexpensive solution. When I release a kit with this feature, I'll include a switch or button that allows you to select the frequency range of the transmitter. This will allow the device to compute the power accurately.

There's a reason why RF devices only have one input - I wanted to see what would happen if I added two inputs to the dummy load. The idea was that it would allow you to use the dummy load with different radios more easily. The problem with the idea is that the leads that go to the unconnected connector act as an antenna. I could privide a loopback connector, but that creates an additional paths for the current to flow, which changes the impedance of the dummy load. It was an interesting experiment, but a poorly conceived idea. The cheapest and easiest way to build a dummy load that works with multiple connectors is to provide a set of adapters.

What's Next?

I don't know if I'm going to release a kit with this feature. For one, I'm not sure if anyone other than me would find it useful. It might be more useful to build a separate device that can be used with any load. With a standalone device and an arbitrary load that you provide, the device could handle a much more powerful transmitter than this dummy load is capable of. If I release a dummy load with this feature, it may use a different microcontroller to compute the wattage and control the OLED display. The Pi Pico is very inexpensive, but it's still overkill for this application. The display is much more expensive to source than the microcontroller, but ironically, it's much less expensive than the old LCD diplays used in 50-year-old calculators.

There's another problem with this design. The Pico needs power. If you're using this in the field, having a seperate power supply may not be convenient. Ideally, the device would be powered by the RF signal itself. I can do that with a voltage rectifier and a capacitor, but that would add to the cost of the kit. I'm not sure if it would be worth it. There are a lot of things to consider. I'll keep you updated on my progress.

73, K7RHY