Solar-powered weather station

One of the students in my IoT Computer Club (here in the UK) is working on a personal project to design and build a solar-powered weather station (that will be located in one of the school's gardens) to monitor and report air temperature, humidity and pressure.

Here are the Objectives and list of items.

The student has split the project into six manageable stages.

  1. Using a solar panel to charge a Li-Ion battery and power a Wemos D1 Mini.
  2. Putting the Wemos D1 Mini into 'deepsleep' to conserve the battery's charge
  3. Switching on/off ancillary devices (e.g. BME280 and ADS1115 modules) to conserve energy
  4. Measuring and reporting air temperature, humidity and pressure and sending this data to the 'Cloud'
  5. Writing a Node-RED flow to process and display the data (extracted from the 'Cloud')
  6. Designing and building an enclosure for the weather station

(Note: The above bullet-points will become hyperlinks as and when the documentation is written.)

So far the student has completed stages 1, 2, 3 and 6 and is working on stages 4 and 5.

I've explained the importance of documenting the work, but like a lot of people... 'writing code and getting something to work is more fun than writing it up'.

Any feedback or encouragement anyone could offer would be appreciated and would probably encourage the student to 'put pen to paper' or whatever the equivalent saying is for an iPad.

Kind regards from David.

1 Like

Sounds Good Dave - i just finished reading an interesting write up someone did using the Sparkfun Thing - key as it has an integrated LiPo battery connector and charing circuitry

I then found another write up that may be useable as it demonstrates appropriate components and how to document a project such as this

Craig

Thanks for the response.
We are using a Wemos D1 Mini Pro (like an ordinary Wemos except you can fit an external anntena for better WiFi range) flashed with ESP Easy firmware and powered from a 2500mAh Li-Ion battery.

Wemos actually have a LiPo daughter board. I have one but I've never got round to trying it yet as all of my sensor platforms so far have been mains powered via USB adaptors.

Some random ideas for encouragement:

  • Get them to sign up to one of the maker sites and to use that to document their progress - getting feedback and encouragement from the whole world can be rather motivating! Parental approval required of course.

  • Get them to break down the documentation like the project and get it on the wall at school where others can see it.

  • Do a few of the boring bits for them and let them just do diagrams.

  • Try to get them using JSDoc when writing code or at least to start with comments outlining the structure before actually coding.

  • Some kind of reward system for producing documentation? Maybe even get parents to agree to provide some rewards?

1 Like

Thanks Julian for your excellent input, which I'll be using.

1 Like

Bullet point 6 (in my orginal posting at the top of this thread) has been updated to show some images of the panels that have been designed using Autodesk Fusion-360.

I've just posted a link (at the top of this thread) to the write-up for... "Objectives and list of items".

I hope you enjoy reading about the first part of this project - hopefully more to come.

I've just posted a link (at the top of this thread) to the write-up for... "Using a solar panel to charge a Li-Ion battery and power a Wemos D1 Mini".

I hope you enjoy reading about Stage 1 of this project - hopefully more to come.

You have probably already seen this small mistake in the link:

"What was really useful was a red LED on the
TP-4056 that indicated the solar-panel was being charged."

Great stuff, I am now looking for a TP-4056 board as I have a few solar panels laying about :slight_smile:

Drat and double drat - I checked his work for punctuation but obviously hadn't put my brain in gear to check the wording made sense.

Thanks for pointing it out.

PS:
The TP-4056 is a great little board in this particular situation.

Although I have yet to find this myself, there has been a lot of talk ever since the little tp4056 boards (just to be clear I am referring to the actual power io boards not the chips themselves) started being produced, about "fake" ones, or units that claim to have voltage/current regulation built in which actually don't.

As the tp4056 is designed for use with Li-Ion batteries it's always best to get them from a reputable supplier, and make sure they come with the DW01 battery protection IC built on the board.

Hi Bob,
That's a very good point you made.
The board mentioned and illustrated in Topic-1 or Stage-1 has indeed got this battery protection device.
The board used in this project has three chips.
The large one on the left-hand-side is the Li-Ion charging controller, the one at bottom-right is a dual channel N-MOSFET and the one at top-right is the DW01V (battery protecion IC).

Fair-weather experiment.

For the last three weeks we have been doing an experiment to see how long the Li-Ion battery will power the Wemos. We started off with a 'deepsleep' time of 20-mins for the first week, then 15-mins and now it's been ticking along at 10-minute intervals for nearly a week.

All the hardware is housed in a cardboard box (as we won't have access to the school's laser cutting machine to make Stage-6 until September) so the project is only put outside when the weather is dry.
We do however keep the project running (24 * 7) even if it is indoors.

Please note: other types of cardboard boxes are available!!!

Always fun to do this type of experimentation. Have you run it with radio only switched off? I forget which mode that is now, but seem to recall significant power savings could also be made in this mode, while still allowing sensor input. Then, every x period, fire up the wifi and send your data packets, back to Node Red naturally. :laughing:

Great fun indeed.

Yes - that's what we have been doing. Putting the Wemos D1 Mini into 'deepsleep' for a long period of time.
We measured the current the Wemos takes during normal operation to be 170mA and when it goes into 'deepsleep' it goes down to 0.2mA. Quite a saving.

We are using ESP Easy as the firmware and have written some 'rules' so when the Wemos wakes up it takes some measurements and send the data via MQTT to Node-RED and optionally also send the information to ThingSpeak.

Stage-2 has been documented.
There is a link at the top of this thread.

1 Like

Such great fun being a student these days. The closest I got was taking the valves out of a television.

I've just posted a link (at the top of this thread) to the write-up for... "Putting the Wemos D1 Mini into 'deepsleep' to conserve the battery's charge".

I hope you enjoy reading about Stage 2 of this project - hopefully more to come.

1 Like

I've just posted a link (at the top of this thread) to the write-up for... " Switching on/off ancillary devices (e.g. BME280 and ADS1115 modules) to conserve energy".

I hope you enjoy reading about Stage 3 of this project - hopefully more to come.

1 Like