After letting the humidity/temperature sensor run on its own battery power with solar charging, I determined I need to have an understanding of how much battery power is being consumed, how much solar power is being generated, and how expensive (power) it is to send a transmission, e.g. 6 transmissions a minute. This is because I found the unit depleted after 24-48 hours and leaving it in the greenhouse during a sunny day did not revive it.
Therefore, I have added power monitoring to the solar panel connection and the connection of the power unit to the Feather (main processing board).
| # | Description |
| 1 | Solar Panel 5v 1.2w |
| 2 | USB / DC / Solar Lithium Ion/Polymer charger – bq24074 |
| 3 | Light sensor 1: right-angle VEML 7700 |
| 4 | Light sensor 2: TSL2591 High Dynamic Range |
| 5 | Battery – 350mAh |
| 6 | Triple Power Monitor |
| 7 | 5 Port “Stemma” [plug] port |
| 8 | All weather humidity & temperature sensor SHT-30 |
| 9 | Adafruit Feather M0 [ARM processor] with RFM95 LoRa Radio – 900MHz – RadioFruit |
Of course, then I began to wonder about how much sun is coming through, so I’m trying two different light sensors that will measure the LUX. While I have a nice IP67 rated case, I need to know just how large a battery and solar panel I need to ensure this unit can make it through several days, so I’m not committing to a case at this time, though I will stuff all these components into one just for testing. I will have my transmissions much less frequent during the night and during days when the fogger would not be needed for cooling, but perhaps just for humidity.
I have achieved, thanks to ChatGPT, confirmation I can access all the sensors and capture their data:19:51:07.595 -> A (Panel→Chg): Vbus=0.616 V, I=0.0 mA, P≈0 mW
19:51:07.595 -> B (Batt↔Chg) : Vbus=0.000 V, I=0.0 mA, P≈0 mW
19:51:07.595 -> C (Chg→Load) : Vbus=4.088 V, I=0.0 mA, P≈0 mW
19:51:07.661 -> SHT-30: T=24.03 °C, RH=59.93 %
19:51:07.894 -> TSL2591: Lux=3.92, Full=88, IR=23, Visible=65
19:51:07.894 -> VEML7700: Lux=15.21
19:51:07.894 -> Feather A7 VBAT: 4.130 V
19:51:07.894 -> ----
Here’s a field test unit as of Friday, August 29, 2025:
Development proceeds. I had a problem with the fogger, when it would halt, the hose would pop off and cause major flooding in the greenhouse!! It took a week to get a replacement part to me which arrived Friday evening at 7:30 PM; I swapped in the replacement and everything looks good now: no more hose popping off at power down.
Here are some progress graphs of the data showing both power generation and consumption and then the humidity when cycling the fogger in a 5 minutes ON, 10 minutes OFF cycle.
It’s looking like solar power is not going to be practical at all… the reduced light from the polycarbonate roofing is working against me and the size needed, e.g. 10 watts per unit, to assure several days of power for rainy cloudy days is causing the costs to skyrockets, e.g. $60 for a 10w solar panel. I’m now considering abandoning solar power and possibly radio communication, I can just provide hard-wired power from within the greenhouse to a small board, e.b. Raspberry Pi Zero 2W, and use WIFI to transmit date and just abandon the entire Lora approach.
Interesting how getting the real data can alter the course of a design. It’s a very fun journey and have ChatGPT to confer with has been critical.
Leave a Reply