Hi everyone,
I’m currently in the early planning stages of a long-term IoT-based monitoring project and would appreciate some advice from the community here. The project will involve remote data logging and wireless communication, ideally using LoRa or Wi-Fi, and I’m looking for a reliable, energy-efficient microcontroller platform that’s well-supported in Australia.
A few questions I’d love input on:
- What microcontroller boards (e.g. ESP32, Raspberry Pi Pico W, Arduino MKR) have you found most reliable in the field?
- Are there any locally available components or modules you’d recommend for reducing lead time?
- Have you had experience with battery management or solar integration for off-grid electronics in the Australian climate?
Looking forward to learning from the experiences of others who’ve worked on similar applications here in Australia.
Thanks in advance!
2 Likes
Hi @Jacob6785295644
Welcome to the forum!
In terms of warranty returns we don’t really see one microcontroller being returned more than another, we will usually go for a Pi Pico for our projects at Core due to their availability to us (currently we have warehoused several thousand across the Pico and Pico2 range and their wireless counterparts). My personal preference is either the ESP32-C3 or ESP32-S3 solely based on having a fair few ready to go at home.
In terms of battery/solar management you have a couple of notable options,
Another notable mention would be Arduino MKR NB 1500 | Buy in Australia | ABX00019 | Core Electronics its got provisions on the board for a sim card so its great for IoT stuff where you’re out of WiFi range and don’t want to set up the infrastructure for LoRa.
2 Likes
Hi Jacob,
A large factor is… what are you used to ? I came from commercial programming so the Raspberry Pi (full size) was pretty much a no-brainer when taking this up as a hobby; because of running linux with wide variety of tools available. It does tend to be overkill for IoT projects, and I now favour ESP32 running ESPHome because there’s a huge number of sensors etc already available, and its “programing” consists mainly of configuration.
If you are familiar with C++ then you will probably favour Arduino and the efficient low-level control it gives.
Re components … I don’t trust my soldering so have been assembling breakout boards. I placed a couple of orders from Adafruit in the early days, but find Core so much more convenient now. Watch out that they stock the same product from different suppliers with different prices.
I am not happy about my aliexpress purchases … the prices are ridiculously low - but too many sellers give misleading descriptions and don’t answer messages. It is really hard to know if you are buying from the manufacturer (eg sonoff) or just another reseller who knows nothing about the product.
solar power
My current project is a hobby (1.8m square) greenhouse which requires solar power. I am using an ESP32-S3, a DFRobot Solar Power Manager 5V module (which does not even monitor the battery voltage), a 6600mAH LiPo battery I had from a previous project, and I bought a cheap solar panel from china (which is not what was described, but good enough).
The graph below shows LiPo battery voltage.
- the first section was a string of sunny days; the power manager (or battery) did not charge over 4.2V. I am using the ESP32’s deep_sleep to shutdown the ESP32 for half hour then wake for 5 minutes.
- the middle section shows thick lines because I managed to stop the deep_sleep - resulting in a lot more sensor readings and power being used 24/7. You can see how much more power was consumed until I realised and fixed my configuration
- third section is back to deep_sleep, but the weather was not so sunny.
My goal was to allow for 3 days of rainy weather before I have to run out to the greenhouse with a powerbank to boost the LiPo’s charge - and the next 3 days (21-23 May) obliged. Next screenshot shows both the battery voltage, and the current being generated by the solar panel…
Didn’t quite get down to 3.4V where my automation on the ESP32 will start sending me alerts or 3.2V where it extends the deep_sleep to 6 hours to conserve the remaining battery.
Even with the solar panel generating < 0.2A on sunny days it took 5 days to bring the battery up to full. I really am happy with that result, especially considering I didn’t even attempt to do the math beforehand
Of course your mileage will be different - there are so many factors.
3 Likes
