Project Caretaker. Part 9. Charging

Since one of our goals is remote control, we should address the power supply issue.

There's no way to expand the battery - the dimensions are already at the minimum possible. So, we need a charging station, and one that the robot can drive onto via video connection. We also need charging process indication and some kind of contact pair. Something similar to the charging stations of robot vacuum cleaners, but made from accessible materials.

I considered magnetic contacts, but due to the apparent complexity of implementation, I settled on spring-loaded ones. I placed pins in the robot's body and plates on the charging station. To save space, springs and indicators are on the station side.

Guided by the same principle of excessive reliability, I purchased 5mm copper rod and 20x20x1.5mm plates. The case was designed around these:

As a power source, I initially planned to use a 5V to 12.6V converter with USB Type-C on board - the specifications promised 3A of current:

But the devil was in the details - 3A was on the input, not the output, and besides that, the board quickly overheats and fails in the absence of active cooling, melting everything around it. From reviews, it was established that the power could be reduced by half by desoldering one of the resistors at the input:

But even this didn't help - the board provided less current and didn't heat up as quickly, but it still reached 70-80 degrees in a minute, which I didn't feel comfortable leaving unattended. In the end, everything came down to a proper external power supply, for which, by happy coincidence, I had a suitable socket connector in my supplies.

The parts fit together quite tightly, but without play. The plates with one spring in the corner on the pin side work perfectly.

Regarding indicators - I had an idea to assemble a circuit with 2 LEDs, but honestly, I didn't quickly figure out the basic setup and schematic, and time was pressing. Instead, I used a volt-ammeter and positioned it on the charging station directly in front of the camera. We'll determine the charging status from its readings. It doesn't show small current, though, so you can either orient by voltage drop or, for significant discharge, by the ammeter.

The final look of the product is as follows:

The station is designed to be this wide specifically to make incorrect docking impossible. The width is a few millimeters less than the distance between the tracks:

And here's the robot standing on the charging station. The camera's focal length doesn't allow seeing everything clearly, but it's sufficient for this kind of indication.

The picture in the stream looks like this (it doesn't depend much on resolution due to focus, but here it's low):

Regarding the charging circuit - inside there's also a 3S BMS board, which controls the charge and ensures equal voltage across the cells. It would be possible to do without it, many charge 3S setups directly, but I prefer the safer option.

Due to the excessive springiness of the springs, placing the robot on the charger is non-trivial - it bounces off and remains without contact, so you have to find the right position. I think a magnetic scheme, like on many wristwatches, would be easier to use. I didn't get around to implementing it, so perhaps the charging station will receive an upgrade in the future.