Finally!
I had some issues with the electronics
LM393 ain’t just plug the wires and we are done. Requires pull up? load? to the output pin. The range of the input can’t be “what ever”. The circuit that finally worked is the following
Notice the voltage divider VR2 for the input signal. The high value coming from the opto-interrupter is almost 3.3V and low 2.4V (maybe) and the LM393 was not triggering. I used that pot to make the signal around 1.2 to 1.6 (or similar) and the IC worked fine. The board with the LM393 and the connectors and all is little bit large. I didn’t cut any excess PCB, felt in vain. The one in the photo is the second version. The first one was more compact, but my soldering is not good yet, so it was messy and not able to locate the error in the transfer from the breadboard. So, I rebuilt it in a full perf board with extra room between the components. It didn’t work, I checked the schematic and found two wrong wires and finally fixed it. Not the most beautiful, but works.
After building the circuit and still wondering why, I decided to have a closer look at the data sheet.
Is this it? Supply voltage in my case 3.3V, this means input max should be 3.3-1.5 or 3.3-2? Makes sense I made input voltage close to 1.2 and worked.
I was waiting also for some RJ11 connectors from Ali, but the connectors that I order were not RJ11!! Were the narrow ones that look like RJ11 for the head phone of the telephone. So I got frustrated and cut off the connectors and screwed both ends to the green terminal block connectors. Not very visible to the next out of focus photo. Camera focused on the anti-bird net and not to the real star…
I also change the funnel to bottle top. Seems to drain the water better than the yogurt cup. The question is when the birds will damage it. Or hot glue failure. Acrylic sides are glued with hot glue and today I find out that they were separated from the glue. Don’t know if the issue is the hot glue or the acrylic it self. At this time I think that I should add a screw to each side to keep the from falling in case of glue failure. People pass by under this location.
The boards were installed in “Station 1”. Too much space. Don’t know maybe at some point I will design a PCB and order from JLCPCP to integrate all components to a single board. But I kinda like the “Soviet” looks of it.
I also I updated the server to store rain height info and instead of average the daily/monthly values, but sum them. Not properly labeled in the app, but the info is there.
Now I am waiting for rain.
