@The_Hedgehog on Project FlowMark
I made a buoy that can float freely down a river. It has four sensors, pH, conductivity, dissolved organic matter, and turbidity. It constantly uses these sensors to take measurements and sends it to a locally hosted Python Flask server tunneled using Ngrok. After the buoy is collected from the water, the server will analyse the data and find out where the water quality started to worsen, therefore finding out where riverine polution might be.
Although the project wasn’t very complicated, I faced many many problems. The most memorable and most fustrating ones:
-
I use a step up module to turn the 4v from my 18650 to 5v for my ESP32 C3 Supermini. But the step up often randomly output over 20v, though it is set for 5v. I later found out that its because a loose connection between my battery and the step up, but the voltage spikes has badly burnt many things in my circuit.
-
The ESP32 C3 Supermini suddenly cannot connect to my computer. I tried holding down boot button, boot and reset, redownload a driver, and trying all the ports on my computer and in Arduino IDE. But it just won’t work. So I changed to a Arduino Nano. But that just made things worse. The Arduino Nano successfully connected to my computer, but it often corrupts recieved UART signals from my cellular module. Later I found out that its because the Arduino Nano’s clock is a bit off for 112500 baud. This made things more complicated because I need to start the UART communication between the Nano and the cellular module at 112500 (since my cellular module is default 112500 when start up), then request the cellular module to turn down to 9600, and then change the Nano UART back to 9600. It did finally work tho.
-
I guess this is the worst nightmare in my project. My project uses a pair of 18650 in parallel, as it can have more power to sustain my buoy for longer. But the cellular module often shut down right after it turns on. It turns out that the celular module can pull up to 2 amps when connecting to the cellular tower and sending messages, which is causing the voltage drop. However, this was weird because the two 18650 cells I had is the Samsung 25R, which has a max discharge of 20 amps. Parallel means that the max discharge become 40 amps, which is wayyyyyyy more than the 2 amps required. So the problem must not be the batteries. Or is it? I suspect that the batteries are fake. They weigh pretty normal, being 44 grams. I checked the imprints on the wrapping, and it did have a legitemate manufacture date code that says someday in 2017 (I forgot which day it was, but it was 2017). And I also took off the battery wrapping, and it did had Samsung factory imprints on the metal. So the battery is good, the cellular module can only take up to 2 amps….. then the problem must be between the cellular module and the battery! And it was. It was the thin jumper wires and breadboard that decreased the voltage when the current increased. Shorter thicker wires solved it.
There was actually much more issues, but these are the most memorable ones for me.
I am glad that my project turns out to be working, becuase i genuinely thought that it would never work when making it.
- 29 devlogs
- 106h build
