Not all of us understand weather mechanisms, how storms develop or what a low pressure front means so having a weather station that measures these parameters and display them as numbers might not be that helpful. However a team from Cornell University designed and built a station that displays a graphic representation of the current weather condition. If its stormy outside you will see a cloud with some raindrops and a lightning, all this on a 8×8 RGB LED matrix.
There are two main modules making this station. The outdoor module who makes all the measurements and then sends the data wirelessly to the indoor module which will display the icon corresponding to the data received. The data transmission is done onÂ 433MhzÂ by two Linx Technologies modules models RXD-433-KH2 and TXD-433-KH2. These modules also take care of data encoding and decoding using an address code so that only your specified modules will receive the transmission.
The outside station has five sensors to measure weather parameters. A solar cell, which also provides power, measures the light intensity and so it can tell if it’s day or night, cloudy or sunny. A LM34 is used as the temperature sensor and outputs a 10mV with every degree F. Rain sensor is from a Hot Wheels Radar Gun, using a Doppler Radar operating at 10.525GHz is sensitive enough to measure velocity of rain drops. A Diy anemometer was the choice as wind speed measurement tool and a capacitor who changes capacitance based on humidity is used for the last measurement.
ATmega644 is the microcontroller gathering data from all the sensors. This is then transmitted as a 10 bits code, each sensor data havingÂ 2 bits in that message. Inside the indoor module we find another ATmega644 who controls the LED matrix. Using PWM to drive the LEDs a very high number of colors is achieved instead of the usual 8 colors.
This project offers a large amount of information and insight into weather monitoring systems. It will require some skills to built it, specially the anemometer, it not as easy as it looks but it’s worth it.
Here is something that could really replace the walking stick. A team from Cornell University developed a navigation system using an ultrasonic sensor that can help you move around without using your sight. The interaction between the obstacle detector and the operator is done usingÂ vibrator motors.
Looking at the block diagram we can distinguish five main areas or modules if you like. We have the haptic interface with the vibrator motors and the ULN2803 IC that drives them, we have the LV-MaxSonar-EZ1 ultrasonic range finder with the bipolar stepper motor and the full bridge driver L298HN, we have the RF transmitter using RCT-433 that sends data wirelessly to a computer for further analysis, we have the ATmega644 microcontroller, the system’s brain and we have the power supply batteries.
The ultrasonic sensor scans the area for obstacle and if it finds any gives out the distance to it. The scanning is done with the help of the stepper motor who rotates the sensor from left to right and back again, covering 270 degrees. Each step rotates the motor 7.5 degrees so it needs 36 steps for a complete sweep. This gives the direction of the obstacle.
Information about the distance to the obstacle and direction is then sent to the ATmega644. After the microcontroller processes the data, it controls the vibrator motors. These motors are spaced 45 degrees apart around the head. The intensity of their vibration varies depending on how close the obstacle is.
The microcontroller also sends out data on RF for further processing on a PC. This data is composed of range information at different angles, stepper motor angle and direction of rotation, vibrator motor status. This data is sent every 45 degrees of the stepper’s rotation.
This is a very documented project with lots of pictures, all schematics and source codes, all the build experience, what things you should be careful about, the problems that were encountered, it’s all there.
When POVs first appeared in advertising panels i used to wonder how are the characters diplayed. Seemed like magic. Later on i found out its because of our eyes’ inertia. Where you needed a large number of LEDs to display a message, now you just need to spin a few LEDs. The rotation speed must be fast enough to display at least 10 frames per second, complex graphics may require a higher value between 15 and 30 fps, movies usually have between 24 and 30 fps.
This project will show you with great details how to build such a device. It is not an easy build, it takes a fair amount of tweaking to get it to work but the results can be spectacular. The developers of this project decided on a modular design, putting an emphasis on interactivity. In the end they came up with a very customizable POV that can display images you upload wirelessly and that you can manage in real time.
The microcontroller used is the Atmega644, leds are driven with the MAX6971 IC and Xbee modules provide the wireless serial communication between the POV and the PC from which you upload the image. A GUI written in Java makes this task easy for you.
The LEDs are place on a different PCB than the rest of electronics, this way it can be easily upgraded to a 3D version. The motor part needs some attention because you will have to carefully balance the LED board, you will deal with a lot of vibrations and the speed of rotation can be quite dangerous. Also there will be some tweaking involved when you will build the brushing system to power up the LEDs.
The motor has a separate power supply so that the main PCB won’t pick up noise. A HALL sensor is used to count the rotations and give the position of the LED board. When real time management of the display is not desired, the POV can use the picture stored in its EEPROM.
In order to power it, you will need a 9v power adapter that can deliver at least 1.1A for the electronics boards. Since the motor is powered from different source, you will need to meet your motor’s requirments. The one used in the project had its own 5V/2.5A power supply.
The cost for building this project is a little higher than 50$ and you can further reduce that if you already have some of the needed parts . All schematics and code are found in the project, below you can see the POV in action.