It seems like I’m really into alliterations.

Not everything went exactly the way I planned it and there are still so many things to do. The project is in such a state that the box and the software are totally usable. My children love it.

It’s been something like two years since most of the work was done. I didn’t have the time or leisure to improve the system anymore. In the last few weeks I wanted to at least complete the blog post series so that other people may build their own Marta.

Continue reading “4 – Forgotten Failures and Fancy Future Features” →

Several hundreds of amperes in a kid’s toy, what could possibly go wrong?

Even if it’s just a few dozens of amps current when accidentally short circuiting a super capacitor, it’s still a little dangerous. Nevertheless, it seemed to be a reasonable idea at the time.

In my idealized world, a super-capacitor should provide enough capacitance to run the machine long enough to safely shutdown the whole system. As soon as the input was disconnected, the Pi was supposed to be running for a few seconds from the power supplied by the capacitor. As the main source is detached physically by a switch, no leakage current could occur afterwards. We’d be save even if the battery was empty.

Continue reading “4.1 – Super-Capacitors + Diodes + Resistors” →

If Tesla powers a car with 7,104 of those batteries, why shouldn’t I use four of them?

Type 18650 batteries can be found in many applications. The huge battery packs used in electric vehicles often consist of thousands of those units. Hook up a few of the cells in series and parallel, et voilà you have your notebook battery.

Wanting to utilize these lithium-ion batteries as a portable power source for Marta, I was planning on increasing the capacity by wiring up four of them in parallel using a battery holder.

Continue reading “4.2 – 18650 type 3.7V Li-Ion Batteries + Charger/Discharger Step-Up Converter + Battery Protection” →

“I’m sorry, my software is perfect. You must be the problem.”

As I have already mentioned, the software was developed to proof that the hardware works and can be used by children. It’s still in an unfinished condition. By the way, I have never ever in my life seen a sufficiently complex piece of software that was ‘finished’. Marta is usable in its current state, not more, not less.

If I had the time, here are a few things I would try to improve or develop:

Continue reading “4.3 – Software Improvements and Ideas” →

I will not touch that PCB design anymore. It works and that’s good. Enclosure improvements are okay.

One could easily hook up an RFID reader to an ESP32, add a battery and speak to the Alexa or Sonos API in order to play a song as soon as a specific tag has been recognized. But that’s not what I had in mind when designing the toy. It was always supposed to be some kind of self-contained device. That means, I would not sacrifice the speaker setup in favor of a bluetooth only solution. We’d trade a little space for a more complex setup and that’s something I don’t like. The device should work all by itself.

But just like the software, the hardware is not completed. It can always be optimized. Here are few ideas I have in mind:

• I added two general purpose connectors on the PCB. They both consist of +5V, +3.3V, GND and 5 unused GPIO pins. It would be nice, if the enclosure had appropriate connectors, too. I ordered two 8 pin GX16 aviation plugs which should work out. The box has enough space left to include these on one of the shorter sites.

Continue reading “4.4 – Hardware Improvements and Ideas” →