This off season, our team is looking into doing various projects. One thing that we would really like to get done is making use of sensors such as encoders, potentiometers and gyroscopes.
Our team has never had much luck with sensors. The extent of our sensor use was a hall effect sensor to measure shooter RPM last year, and a limit switch to tell us when our catapult was retracted. We tried to use Greyhill 63R encoders last year, but with no luck. (1114 and 2056 tried to help us get them working at GTR East last year, with no avail).
This off season, we have had some luck with the old KOP Gyroscope/Accelerometer sensor to detect our robots rotation.
Our main question is, what kind of sensors do you recommend? We’ve browsed many websites looking for encoders, potentiometers and gyro’s but we are lost on which ones work the best.
If you could tell us what kind of encoders you use on drive trains and shooters, the potentiometers you use and the gyro’s you use, why you use them, and how they have worked for you, we would really appreciate it.
For encoders, this year 449 used a US digital H6 Ball bearing encoder (512 CPR) for our drive train, and it worked magically. We like it because it’s a heavy duty encoder that is far more robust than the KOP E4P, and it’s pretty easy to install.
Last for our last year's robot, our team used both a potentiometer and the KOP accelerometer on our robot. For potentiometers you can basically use any run of the mill 1-10K pot, as long as you can find a way to attach a shaft to it. Also, the KOP gyros and sensors aren't terrible. If you want something better then you could try out a COTS IMU (I Know sparkfun has a few).
Also as a note for analog sensors, you might also want to experiment with various filtering methods to reduce signal noise. Both hardware methods (ie simple low pass filter) and software (LabView is great for this).
For quick and dirty sensing our team has had a good experience with anything VEX, largely because of their robustness and ease of mounting and wiring. The ultrasonic rangefinder, limit switch, and bumper switch all required very little work (assuming things like female-female PWMs were on hand).
As mentioned earlier, any 1K to 10K-ohm potentiometer should work well. You can use a shaft coupling to attach it to a shaft.
The main question comes down to what do you want to sense, and then how do you do that.
We’ve had good luck with the KOP Gyro/Accel, but are looking at alternatives due to cost.
For quadrature encoders, we typically use the US Digital ones from andymark, but would not be averse to finding alternatives due to how fragile the connector is. Absolute encoders have varied. Below are two other sensors we used this year and where we bought them. We also use SparkFun as a source.
That’s the key question. What do you want to sense?
For our swerve drive, we need to know absolute angle: 360 degrees and stop-less. We use BI Technologies Magnepot Hall Effect Potentiometer Part # 6127V1A360L.5FS to do this. The analog sensor works very well and is ~$11 ea.
Axis M11 Camerax2
Vision tracking and a ball pickup camera.
Distance/angle to Target using binocular vision. Ball Pickup is something we’re looking at adding now that we have an example to work from. We used this system as (at the time) it provided better accuracy & noise rejection over using one.
Hall effect x1
Shooter feedback
Correct
**Proximity Sensor x3 **
Ball sensors?
Correct. We used this for managing the balls in the robot. We maintained positive control & position knowlege over all three at all times through the use of a state machine.
**Light/Dark sensor x1 **
?
Used for telling us when our turrent was in the home position.
IR Prox Switch x1 (from LogoMotion)
?
We used this at the front of the ball intake to let us know when a new ball was coming in. The proximity sensors were further up the system, and with the three ball limit we wanted to ensure we never possessed a fourth.
Limit Switch x1
Turret soft stop
We had a “spear” that would come down for lowering the bridge. It was used as a softstop for that.
Encoder x2
Drivetrain
Correct.
**Absolute Encoder x1 (analog) **
Turret feedback
Correct, although we would loop it and had to add code which would note this. Our turret had a range of about 3 full encoder revolutions.