Windmeter with smartphone sensor

[Erwin_Meyvaert]

I have asked ChatGPT this, but I do not understand how to write this correctly in mit app inventor, especially I do not know how to place the formulas

Here is a detailed step-by-step guide to writing an MIT App Inventor program that measures wind speed and direction using a magnetometer sensor:

Open MIT App Inventor and create a new project.
Drag a "Screen" component onto the screen and give it an appropriate name, for example "Measure Wind Speed".
Drag two "Label" components onto the screen and place them in a convenient location. Name one label "Wind Speed" and the other label "Wind Direction".
Drag a "MagnetometerSensor" component onto the screen and name it "magnetometerSensor".
Go to the "Blocks Editor" and drag a "When Screen1.Initialize" block onto the stage. This block is executed when the screen is initialized.
Within the "When Screen1.Initialize" block, add a "Set" block to set the magnetometer sensor values to zero (0).
Drag a "When MagnetometerSensor.MagneticFieldChanged" block onto the Stage. This block is executed when a change in the magnetic field is detected.
Inside the "When MagnetometerSensor.MagneticFieldChanged" block, add a "Set" block to store the values of the magnetometer sensor in three variables: "x", "y" and "z".
Drag a "Calculate" block to the work area and add the formula to calculate the total magnetic force: magnetic force = √(x^2 + y^2 + z^2).
Inside the "Calculate" block we just created, add a "Set" block to store the value of the total magnetic force in a variable, for example "MagneticForce".
Drag another "Calculate" block onto the work area and add the formula to calculate the wind speed: wind speed = magnetic force * magnet radius * pi / 30. You will need to enter the magnet radius you used to place the magnet on top of the phone. to place.
Inside the "Calculate" block we just created, add a "Set" block to store the wind speed value in a variable, e.g. "wind speed".
Drag another "Calculate" block onto the work area and add the formula to calculate the wind direction: wind direction = arctan2(y, x) * 180 / pi. Here the "arctan2" function is used to determine the angle between the y-axis and x-axis, and the result is converted to degrees.
Inside the "Calculate" block we just created, add a "Set" block to store the wind direction value in a variable, e.g. "wind direction
Drag a "Label1.Text" block onto the Stage and place it inside the "When MagnetometerSensor.MagneticFieldChanged" block. Set the value of the "Text" property of this block to "Wind speed: " + wind speed + " m/s". This ensures that the wind speed is displayed in the corresponding label on the screen.
Drag a "Label2.Text" block onto the Stage and place it inside the "When MagnetometerSensor.MagneticFieldChanged" block. Set the value of the "Text" property of this block to "Wind direction: " + wind direction + " °". This ensures that the wind direction is shown in the corresponding label on the screen.
Add extra functions if necessary, such as a button to stop the program or a timer to repeat the measurements at set times.
Test the program on your own mobile device and make sure it works correctly.
It is important to note that the accuracy of the measurements depends on the quality of the magnetometer sensor and the distance between the magnet and the sensor. To improve accuracy, it may be necessary to calibrate the magnet and/or sensor. This can be done using a special calibration app or program.

It is also important to observe safety regulations when working with magnets. Make sure that the magnet is not too strong and that it does not come close to sensitive electronics or magnetic storage media such as credit cards and hard drives.

[ABG]

@Erwin_Meyvaert

And finally, the wind direction formula can be calculated by taking the arcsin of the ratio of Y and X, followed by a conversion to degrees.
But i don't know how to translate this in mit app inventor

This part sounds like it needs the orientation sensor?

[ABG]

The ChatGPT paragraph sounds very realistic.
(I voted for a Congressman like that recently, to my regret.)

But I don't see how you could get wind direction without a weathervane in addition to the speed whirlygig.

[Erwin_Meyvaert]

I had this in mind and now found it on the internet

Ontwerp je eigen chip - PDF Free Download

It's about halfway through the magazine
It is in Dutch maybe you can translate it with google
It's a hardware project but it's the principle that counts that might be useful for the project
But first start with an ordinary anemometer

[Erwin_Meyvaert]

Can someone translate ChatGPT's story into mit app inventor please , or partially especially the formulas and such
Seems like a great and educational project to me

[ABG]

To measure wind direction with your phone:

Place the phone in a plastic sandwich bag to protect it from the rain, opening downwards.
Staple the bag at the bottom to stop the phone from falling out.
Insert a long pencil or dowel into the bag to the side of the phone, to form a handheld flag.
Raise your flag, dowel perfectly vertical, and let the wind turn your bag flag freely.
Set your phone to record its Orientation Sensor readings along with the Clock.SystemTime for each reading into a global table.

At end of run, convert the global table to CSV and save it to a File, with an appropriately formatted file name built from the first reading timestamp.

[Erwin_Meyvaert]

Can someone help me further please

[Patryk_F]

I think we've all lost the thread. What do you really need?

[Erwin_Meyvaert]

A program in mit app inventor to measure the wind speed and maybe the anemometer (option)
maybe i will start with chatGPT and when i get a start i will come back for help if needed
Thank you all for the information provided

[ABG]

I already posted code to show the rotation rate by counting peaks.

That is a fudge factor away from wind speed.

[Erwin_Meyvaert]

Thank you very much for this info

I have this but the smooth and clear button don't work according to me
Can you perhaps only indicate the speed in km/h for example

[ABG]

I thought further on the wind direction problem.

The hardware you mentioned did indeed have two moving parts,

• a spinning part to measure wind speed by spinning fast in string wind, and
• a slow moving off center paddle to align with the wind direction.

I assume both parts have magnets in them, and the magnetic field detected by the phone is the sum of the magnetic fields induced by the two moving parts.

The spinning part will add a repeating wave pattern to the 4 directional magnetic field strengths (x,y,z,total) if you map them over time.

The paddle induces a slow moving change to the ratios of the magnetic forces in the three directions x,y,z, depending on how you hold the phone (Statue of Liberty torch vs motor boat engine.)

There is precedent for detection of external magnetic field direction in the AI2 Gallery. They are called Compass apps. I have not searched for them recently, I leave that to you.

Be aware of the existence of a very handy trig function atan2(x,y) in the math blocks section. It is superior to the other functions arcsin, arctan, arccos, in that it avoids the divide by zero problem when getting angles from the x axis to points at (x,y).

Also, be aware that by holding the phone up like a torch, the new directions will change from (x,y,z) to (x,z,y) (depending on how you hold your phone), so the angle would be atan2(xStrength,zStrength) in the Magnetic Field Sensor.

[ABG]

@Erwin_Meyvaert

I have this but the smooth and clear button don't work according to me

(edit - Maybe the app is using the wrong axis for that orientation. Instead of xStrength, use yStrength or zStrength and check the graph for wiggles when you spin the magnet.)

Please download and post each of those event block(s)/procedures here ...
P.S. These blocks can be dragged directly into your Blocks Editor workspace.

See Download Block Images for a demo.

@Erwin_Meyvaert

Can you perhaps only indicate the speed in km/h for example

Translating rotational speed to km/h requires calibration, perhaps by having some one drive you around on a calm day at set speeds (km/h) and recording the rotational speeds at different auto speeds (hold the phone out an open car window and don't drop it.)

Do I need to explain the arithmetic to derive the fudge factor from the calibration run?

[ABG]

Here is a separate app to play with the angles and guess the spin rate without a graph.

Data is collected into a global table, for you to export to a File if you want.

magnetic_spin.aia (6.3 KB)

Feel free to customize it once you have completed your manual calibration of frequency ve land speed.

P.S. I hope the magnetic strength of your rotor exceeds the local magnetic field of your car.

[Erwin_Meyvaert]

thanks for all the info, to be continued

[Erwin_Meyvaert]

I am still looking for a program, I asked ChatGPT and he forwarded it to me

• First you need to go to the MIT App Inventor website and create a new project. You can give your project a name, for example "Anemometer".
• Then you need to go to the "Designer" tab and add a new screen. You can give your screen a name, for example "Screen1".
• On the "Designer" tab, you need to go to the "Palette" and drag and drop the following components onto your screen:
  • A "MagneticFieldSensor" component from the "Sensors" category. This is the component that can measure the strength of the magnetic field in three directions (x, y, z). You do not need to change anything in the properties of this component.
  • A "Label" component from the "User Interface" category. This is the component that will show the speed on the screen. You can change the text of this component to something like "Speed: 0 m/s". You can also adjust the size, font and color of this component to your preference.
• Next you need to go to the "Blocks" tab and write the logic of your program using the blocks. You can find the blocks in the "Blocks" toolbar on the left side of the screen.
• The explanation of each block is as follows:
  • The first block is a "when Screen1.Initialize" block. This block is executed when the screen is opened. In this block you set the "Enabled" property of the "MagneticFieldSensor1" component to true using a "set MagneticFieldSensor1.Enabled to" block and a "true" block.
  • The second block is a "when MagneticFieldSensor1.MagneticChanged" block. This block is executed when the "MagneticFieldSensor1" component detects a change in the magnetic field. In this block you call a procedure that calculates the speed based on the x-strength of the magnetic field. The procedure is called "calculateSpeed" and has one parameter called "x". You use a "call calculateSpeed" block and a "get MagneticFieldSensor1.XStrength" block to call this procedure.
  • The third block is a "to calculateSpeed x" block. This is the definition of the procedure that calculates the speed. This procedure has one parameter called "x", which is the x-strength of the magnetic field. In this procedure you do the following:
    • You create two variables, called "lastTime" and "penultimateTime". These variables store the time of the last and penultimate passage of a magnet past the sensor, respectively. You use two "initialize global name to" blocks to create these variables and initialize them to 0.
    • You check whether the x-strength is greater than a certain threshold value, for example 20. This means that a magnet has passed the sensor. You use an "if" block and a ">" block to do this check.
    • If so, do the following:
      • You set the value of the variable "penultimateTime" equal to the value of the variable "lastTime". You use a "set global name to" block and a "get global name" block to do this.
      • You set the value of the variable "lastTime" equal to the current time in milliseconds. You use another "set global name to" block and a "Clock1.Now" block to do this.
      • You calculate the time difference between the last and penultimate passage of a magnet. You use a "-" block and two "get global name" blocks to do this.
      • You check whether the time difference is greater than 0. This means that there has been a valid rotation. You use another "if" block and a ">" block to do this check.
      • If so, do the following:
        • You calculate the rotation speed in rotations per minute (RPM) using the formula: rotation speed = 1000 / (time difference between last and penultimate pass). You use a "/" block, a "1000" block and a "-" block to do this.
        • You calculate the speed in m/s using the formula: speed = rotational speed * circumference / 60. Here the circumference is equal to 2 * pi * radius, where the radius is the distance between the center of the anemometer and the center of a cup. You can adjust the radius to the actual value of your anemometer, for example 0.05 m. You use a "*" block, a "/" block, a "2" block, a "pi" block, a "0.05" block and a "60" block to do this.
        • You set the text of the "Label1" component equal to the speed, possibly with an additional unit. You use a "set Label1.Text to" block and a "join" block to do this. You can also use a "round" block to round the speed to, for example, two decimal places.

Is there any truth in that??????? Before I investigate further to program it in the app inventor, someone can sometimes check whether this is correct
Best regards

[ABG]

ChatGPT is a politician that tells you what you want to hear.

You still have to think and test.

[Erwin_Meyvaert]

It's huge that he responds like this to a question, I know he sometimes makes mistakes, which is why I ask whether there is anything useful in his explanation.

[ABG]

@Erwin_Meyvaert

why I ask whether there is anything useful in his explanation.

It is beneficial to your understanding of the problem for you to follow the detailed instructions and to see where they go wrong.

Correcting other people's code on this board can be enlightening to both sides of the conversation.

ChatGPT lacks gratitude, though.

[Patryk_F]

For chatgpt to answer correctly, you need to ask your question very carefully and wisely. Too general questions will bring bad answers. You also need to have some knowledge of the subject to spot obvious mistakes.