Favorite Sensors

[MaxMax161]

This season I discovered having good sensors can make all the difference in control systems. My team switched from using the KOP US digital encoders on our shooter wheels (what we used last year) to Greyhill 63R encoders link. These new encoders were easy to use, reliable and made life so much easier and I was wondering what other sensors teams have fallen in love with. What's that gyro, accelerometer, pot, gear tooth counter, limit switches, photogate or encoder that has stolen your heart? Please post part numbers if you have them.

[Kevin Ray]

The sensor that makes us very happy was introduced to us at an Atlanta CMP by another team (forgot which). We've used them every year since. They're a bit expensive but have phenominal utility on a robot. This http://celesco.com/stringpots/ is one of the companies producing these string pots. I don't know how we could have done some of the things we did without them--well I do, but it would have been much more difficult.

[Gdeaver]

For a 360 degree no dead band angular position sensor, I like the cost and performance of the TT Optek 6127V1a360L.5FS. It has worked very well for our swerve steering position sensing. For a quadrature encoder we are looking at the CUI AMT102-V. You can set the CPR by dip switches. So one unit can give 32 to 1024 CPR. We made a mount for it out of PVC with 2 ball bearings and a 1/4" shaft. For a tach the Melexis US2882 Hall effect bipolar latching switch has worked well for many teams. We are exploring the use of a Invensense MPU-6050 3 axis gyro and accelerometer this summer. One of the most stable gyro's that is affordable I have looked at yet.

[Hugh Meyer]

For end of travel, home, or just sensing something moving in and out of position a vane or slot sensor works nice.

http://www.digikey.com/product-searc...819z&x=10&y=13

See attached photo for application.

-Hugh

[Phalanx]

I've been taken by the USDigital H-5 series of shaft encoders.
http://www.usdigital.com/products/en...otary/shaft/H5

I started using them several years ago, when I couldn't get the Greyhill ones fast enough (EVERYONE was out of stock at the time, Mouser, Digikey, Newark).

Turns out, I like them better than the Greyhill 63R because....
1) 10K RPM versus 5K. (our shooter motors ran over 6K RPM)
2) Locking, captured signal cable, doesn't fall out, nor likely to break off the header pins.
3) Discount for FRC Teams.

[ekapalka]

Originally Posted by Phalanx

USDigital H-5 series of shaft encoders

My team (actually just myself) has been reluctant to buy any non-analog sensors other than the encoders supplied by AndyMark because we don't know if we'll be capable of writing our own classes. The encoder you linked looks stellar. Was it much of a pain to program (or could you use the classes supplied in WPILIB)? My personal favourite sensor is the USDigital MA3 Absolute Magnetic Shaft Encoder (analog option!). I wish we could get a Celesco Linear Potentiometer, but they apparently don't have a distributor in our area (Montana) and won't mail us one :/ I have no idea what for, but I'm sure we could think of something

[Phalanx]

Originally Posted by ekapalka

Was it much of a pain to program (or could you use the classes supplied in WPILIB)?

They work just fine using the standard WPILIB encoder methods. Nothing special is required.

[Mr. Lim]

For a 1-turn pot, I always use a Bourns 6639 Series Potentiometer.

They're pretty cheap, and they're continuous so you won't destroy them if you drive them past their extents.

I will never use a fixed 1-turn pot ever again.

We've had success with LV-MaxSonar-EZ Ultrasonic Rangefinders as well. They're easy to work with, and give reliable reads in the right conditions with a simple median filter.

When it comes to optical sensors, Banner's WorldBeam QS18 Series have never let me down. We like to use a visible light one, so we know whether the sensor is working, and what it's pointing at. They've held up in high-speed applications too, as we used one of these to measure our shooter speed by colouring parts of it black and white.

[Garrett.d.w]

My favorite sensor, hands down, is the lowly limit switch. You can do just about anything with these things. That and they are cheap, tough, little buggers.

My second favorite is a broken beam sensor. I don't know the make or model of the ones that we use. They are easy to install as well as easy to program. They also glow red in the dark, which you know, is kinda important

You have to be careful with the broken beam sensors, however. If you order some, make sure they use LEDs for the light rather than a laser. Also, you get improved reliability if they are in an environment where they are not exposed to too much direct light.

[AllenGregoryIV]

Originally Posted by Mr. Lim

For a 1-turn pot, I always use a Bourns 6639 Series Potentiometer.

They're pretty cheap, and they're continuous so you won't destroy them if you drive them past their extents.

I will never use a fixed 1-turn pot ever again.

Thanks for this, I had been looking for something like this for a while.

[Kauaibots2465jm]

Aloha,

Our team, the Kauaibots, has started using the SHARP IR Proximity Sensors that you can buy from Sparkfun.com. They come in various ranges like 3 to 30 cm and 10 to 80 cm. We found that if you bounce the IR beam off of some of the reflective tape provided in the KOP they provide easy, reproducible results. This year we used one to measure the angle of our Frisbee shooter and the height of our climbing hooks. The reflective tape appears to be as important as the sensor.

We have tried to use the accelerometer/gyro boards included with the KOP with poor results for our key desired feature: Field-oriented drive. One of our mentors made our own Inertial Measurement Unit (IMU) circuit board which we used this past year with a lot of success on our field oriented drive with swerve drive. In other words our omni-directional drive robot will go in the direction that the joystick is pushed no matter what the robot's orientation is. This makes our robot much easier to drive. We were encouraged to sell them as a fund raiser for our team so we had a batch made. Now we are looking for about 5 teams to be beta testers. We will provide the board, software, and instruction set for a price of $65 to beta testers. This is below our cost.

There were two main goals for this project. First we wanted something that would work reliably, with yaw rate drift of less than 1 degree/game (which was the part that required the complex algorithms), and secondly we wanted something easy to work with so that our students could do the programming (so it was designed to be programmable with the Arduino IDE, and have also provided source code showing how to access it from the cRIO). We've completed the work and think this could be helpful for other teams.

Here are some of the specs for what we will market as the iNAV9:

iNAV9 Technical Specifications

- Small circuit board (2.5" x 2")
- 3 axis accelerometer and gyro (Invensense MPU-6050)
- On-board Auto-calibration and Digital Motion Processing algorithms - yaw drift of only .15 degree/minute (max of about .5 degrees/per game) means no need to stop/reset in the middle of the match.
- 3-axis magnetometer (Honeywell HMC5883L) (Compass for reading initial alignment before match starts; after match starts the readings are not reliable due to electromagnetic interference). This allows the robot to start out not pointing at the head of the field, and still be usable for field-oriented drive.
- On-board microcontroller (Atmega328) which sends fused yaw/pitch/roll and compass direction to cRio via RS-232 cable. Update rate can be as high as 200Hz, we found 100Hz works great.
- Arduino-compatible and open-source board firmware so anyone can re-program the board using the Arduino IDE.
- RS-232 cable for connection to cRio.
- JST connector for direct connection to 12V unregulated supply from the Power Distribution Board.
- Example code for interfacing to the board from the cRIO via WPI Library.
Instruction set for mounting, checking , calibration and programming.

If you are interested in Beta testing please email us at james@Omaotech.com.