I aim to set up a small system to harvest energy from the constant Wellington wind, using only recycled motors. Currently I intend on using this energy to power a small greenhouse, and as an emergency backup source of electricity.
Ultimately, this project will create a system for safely harvesting and utilising wind energy around my own home.
So, I haven't updated here in a while. I haven't really had any spare time, so this log will cover all the progress I have made in about a month.
First up, I put together a PCB for mounting all the anemomenter components. This is shown below, and the circuit for wind speed measurement is yet to be tested, hopefully I can test that in the next couple of weeks.
I have uploaded the KiCAD project for this board to the project files. Essentially a TC4056A LiPo charger module is used to take power from a small solar panel, and manage the charging of a small LiPo cell. This is obviously not optimal, as no maximum power point tracking will be performed, so likely output will be abysmal. However, this is an easy initial testing stage. Battery voltage is then converted to 3.3v for an ESP32 module. Below this is the circuit for interfacing with the anemometer that I detailed in my last log.
My plan was to test the power system for the anemometer, but I realised that my setup as it is won't cut it. I have a small 1s LiPo that I want to use (For ease of charging), however what I failed to realise was that the only buck converter I have on hand requires a minimum input of 4.5v when stepping down to 3.3v. As a single cell LiPo can supply 4.2v at it's best, this didn't seem like such a good solution anymore. I have ordered a small buck/boost converter that should be able to be dropped into the PCB I have made.
I believe that once I can get 3.3v out from the battery, regardless of the battery voltage, I should be good to start designing a housing. Before then, I need to complete testing on the actual windspeed sensing capabilities. Externally powering the module, and writing up some code to read the anemometer should be enough for a proof of concept.
Secondly, I have done some thinking, and based on the area around my house, a horizontal axis wind turbine may be better than the vertical axis I had in mind initially. I have started sketching up some designs for a gearing system to mount to the scooter motors, and done some sketches for how I might setup my nacelle.
Based on this, I will probably need to add some form of windvane to the anemometer setup. This can be a next step after getting the wind speed data logging working.
Overall, a pretty useless progress update really. I have a university holiday coming up soon, so I am hoping to get stuck in and finish some things here.
I have spent the past couple of days setting up the circuitry to go inside the anemometer design I posted previously. I have setup the following circuit:
U1A and U1B form the infrared receiver/emitter pair of the TCRT5000 sensor. U1A is a phototransistor, so will pull the inverting input of the comparator towards GND when IR light is present on the sensor. R_POT should be adjusted such that the inverting input (pin 2) is lower than the non-inverting input (pin 3) when the sensor is underneath the white section of the headset. Conversely, pin 2 should be a higher voltage than pin 3 when the sensor is beneath the black component of the headset.
As the comparator output is connected through a pull-up resistor, the system will output 3.3v for most of the cycle, and will output a low pulse as the black component passes over the sensor.
In order to verify the integrity of the pulse, I plugged the system into my incredibly dodgy $30 oscilloscope. The results are shown below:
Falling edge of signal (above). It is clear that the signal falls over a period of about 50uS.
Rising edge of signal (above). This also occurs over the period of about 50uS, though this image is not as clear.
Both edges of the low pulse seem to be relatively clean, and should be suitable to use as an input to the microcontroller interrupt pin. The sensor does not appear to bounce, transitioning cleanly and consistently between states.
From here, I will solder up a test control board, and create a mounting system for the circuitry and headpiece. I will update here once I have this, and hopefully have the headpiece measuring windspeed.
I have been using the time in lockdown as an opportunity to develop a wireless anemometer. The intention is to have an anemometer that can be mounted to a pole, charge an internal battery via small solar panel, and transmit wind data to a server. From here, I can develop a frontend that allows me to view the wind data from the intended turbine location.
Ideally, with enough data, I can determine the optimal location for my turbine. It will also just be cool to have realtime local wind data.
So far, I have created a headpiece for the anemometer. The top assembly clamps onto a bearing, which is in turn bolted to the bottom piece. This means the headpiece rotates freely about the base. The top piece has a small black section of plastic on the inside, exposed to an IR sensor mounted to the base piece. Using a TCRT5000 IR sensor, and an LM393 comparator, this sensor can be used to generate a pulse as the black section of plastic passes over the sensor. Ideally I would use a hall effect sensor and a magnet, but for the time being I need to work with the components I have. Below is an image of the headpiece as it currently stands.
Files for the anemometer can be found in this OnShape document. This is a working document, and will be modified as I progress.
There are a couple of issues to address with the current model:
Bolts that clamp to the bearing should be shifted around the headpiece 60 degreses, so as to be out of the way of the cups.
Cup mounting mechanism needs to be tighter, as some of the cups are not mounted securely.
From here, the next step is to develop the circuitry to detect rotation, connect it to the microcontroller, and use this to calculate windspeed. I'll check back in once I have this working.
This site seems like as good a place as any to log my progress on this project, so I will do my best to upload any progress files, and log any advancements here.
As it stands, I am still in the very early stages of this project. I have scrapped a single washing machine motor from a machine my neighbours were throwing out, and was very kindly given 6 Jump scooter motors from the amazing people over at Uber. The first step with these motors will be to analyse how they output power. It will likely be important to know how power output is related to rotation speed, as this will determine whether I can get away with a direct drive system, or whether I will need to design some sort of gear system. I will also need to analyse the IV output graph, and how it changes with generator speed as well. Ideally this analysis will allow me to determine maximum power output conditions, and develop an electronic system to control the generation.
Additionally, it is important to have an idea of wind speeds in the area of the turbine, so that I can ensure the turbine does not operate in dangerous conditions. This will be my first project, building an easily replicated anemometer to get an idea of how the wind behaves around my house. This should also help with choosing a location for the turbines.