June 13, 2004, Undergraduate research for the Espinosa Group, Mechanical Engineering Dept., Northwestern University
MEMS Gyroscope test setup
MEMS Gyroscope Project
Updated: June 13, 2004
Goal: Create a lab that shows how a MEMS gyroscope can sense position of a robot.
Progress:
Niel’s method differs from the final goal of the project in that he didn’t implement the gyroscope onto the robot or established a feedback loop to take advantage of the information gained by the sensor. While recreating what Niel did, I have been considering how to implement the robot as well. The robot itself is run by 2 modified servos and a stamp board. In order to be successful in this lab, a few more components are needed; The obvious one is the MEMS gyroscope which also needs an A/D converter so it’s output signal is recognizable to the stamp processor. Niel also used a timer to run the servo, however from my reading about the stamp board and the servos, it is not apparent why he needed this. The stamp board should be able to run the servos without an external timer. This makes the circuitry easier because when building the robot, two timers would be needed one for each servo. I believe the only outside component needed besides the sensor is its A/D converter. I will attempt to set up the system without the timer as soon as possible.
To date, I have completed the circuitry to spin the robot arm and send signal from the sensor to the serial cable. This has included wireing and testing the outputs of the timer and A/D circuits. To finish the circuitry and hardware components of the robot I would like to solder the gyroscope, A/D converter, and any other necessary external componant onto a board with one power input, one ground, and an easy plug to connect to the stamp board. At that point it will be simple to switch from testing the gyroscope on a rotating arm to implementing it on the robot itself. Then at any point the project can switch from writing test code to writing robot commands.
Niel used two languages and two programs to run his tests: BASICStamp to run the stamp board/servos, and LabView to recieve data from the sensor. This may pose a problem when implementing the sensor in a feedback loop. Instead, all the scripts for the robot will be writen in BASICStamp and run natively on the stamp board. I do not think implementing the MEMS gyroscope is impossible because a variety of sensors are available commercially for this robot with much literature on them. The only thing to mesh this sensor in with the current hardware is to make sure the output is something that it can understand and can be programmed to interpret. The easiest way to do this would be to get the A/D converter’s output to be as close to the output of the commercially available sensors. This way the programing should be much easier.
Next Steps:
Build robot frame
Control servos from stamp board directly
Soulder MEMS gyroscope and A/D converter onto a chip
MEMS Gyroscope Project
June 13, 2004, Undergraduate research for the Espinosa Group, Mechanical Engineering Dept., Northwestern University
MEMS Gyroscope test setup
MEMS Gyroscope Project
Updated: June 13, 2004
Goal: Create a lab that shows how a MEMS gyroscope can sense position of a robot.
Progress:
Niel’s method differs from the final goal of the project in that he didn’t implement the gyroscope onto the robot or established a feedback loop to take advantage of the information gained by the sensor. While recreating what Niel did, I have been considering how to implement the robot as well. The robot itself is run by 2 modified servos and a stamp board. In order to be successful in this lab, a few more components are needed; The obvious one is the MEMS gyroscope which also needs an A/D converter so it’s output signal is recognizable to the stamp processor. Niel also used a timer to run the servo, however from my reading about the stamp board and the servos, it is not apparent why he needed this. The stamp board should be able to run the servos without an external timer. This makes the circuitry easier because when building the robot, two timers would be needed one for each servo. I believe the only outside component needed besides the sensor is its A/D converter. I will attempt to set up the system without the timer as soon as possible.
To date, I have completed the circuitry to spin the robot arm and send signal from the sensor to the serial cable. This has included wireing and testing the outputs of the timer and A/D circuits. To finish the circuitry and hardware components of the robot I would like to solder the gyroscope, A/D converter, and any other necessary external componant onto a board with one power input, one ground, and an easy plug to connect to the stamp board. At that point it will be simple to switch from testing the gyroscope on a rotating arm to implementing it on the robot itself. Then at any point the project can switch from writing test code to writing robot commands.
Niel used two languages and two programs to run his tests: BASICStamp to run the stamp board/servos, and LabView to recieve data from the sensor. This may pose a problem when implementing the sensor in a feedback loop. Instead, all the scripts for the robot will be writen in BASICStamp and run natively on the stamp board. I do not think implementing the MEMS gyroscope is impossible because a variety of sensors are available commercially for this robot with much literature on them. The only thing to mesh this sensor in with the current hardware is to make sure the output is something that it can understand and can be programmed to interpret. The easiest way to do this would be to get the A/D converter’s output to be as close to the output of the commercially available sensors. This way the programing should be much easier.
Next Steps: