Compass/direction sensor

[Frank(Aflak)]

We would like to get our hand on two compasses that will plug into the robot controller and the operator interface. We don't really want to use the angular motion sensor, all we want to know is the anlge between whatever way the OI is facing and whatever way the robot is facing. I'm gonna go hit up google, but if anyone knows a good supplier offhand please respond.

thanks.

edit:

I found this and this but I was kinda hoping to find one ready to plug into the controller and return a number between 0-360 or 720 or whatever.

HEY! word! look at this baby!linky!

[Dave Flowerday]

Note that you're not allowed to hook custom electronics up to the OI...

Also, our team looked extensively at several magnetometers and found that they are essentially worthless when placed inside the robot.

[Ian W.]

but could it be considered an analog input? sort of like a pot, if it returns the correct voltage?

plus, would one of those crazy "R2 Logic Ladders" (or whatever it is when you make an analog input into many digital ones) be illegal then as well?

[Al Skierkiewicz]

Ian,
As Dave said above we did some rather extensive research into direction finding and compass usage. Besides the fact that they are somewhat expensive, they cannot distinguish the earth's magnetic field when placed within the robot frame with all the DC(magnetic) motors and various metals in the construction. There is also so many variables when used inside buildings due to electric fields and metals in the roof, walls and floors. Although they look very good on the surface they became to difficult to implement.
Sorry.
If you are intending on using a feedback system from the OI to the robot, teams have been using the gyro for years to do that so there is a lot of info around. Remember that if the sensors are initialized correctly, they should track fairly well over a two minute match.

[Adam Y.]

As Dave said above we did some rather extensive research into direction finding and compass usage. Besides the fact that they are somewhat expensive, they cannot distinguish the earth's magnetic field when placed within the robot frame with all the DC(magnetic) motors and various metals in the construction. There is also so many variables when used inside buildings due to electric fields and metals in the roof, walls and floors. Although they look very good on the surface they became to difficult to implement.

How bad was the interfernce with the compass?? Which compasses did you try using?? Would having a totally plastic robot fix the problem??? Im asking these questions because Im building robots for fun and it is cheaper for me to use plastics. Ill see what I can find out about them.

[Bduggan04]

Originally posted by wysiswyg
How bad was the interfernce with the compass?? Which compasses did you try using?? Would having a totally plastic robot fix the problem??? Im asking these questions because Im building robots for fun and it is cheaper for me to use plastics. Ill see what I can find out about them.

Only if you had plastic wire. The PWM controls work with pulses which causes a constantly changing magnetic field. And just so you know, the compasses sold through digikey are over $300, far surpassing the custom electronics budget.

[Adam Y.]

Only if you had plastic wire. The PWM controls work with pulses which causes a constantly changing magnetic field. And just so you know, the compasses sold through digikey are over $300, far surpassing the custom electronics budget

By George I have figured out the problem. Wrap the wires around each other. The magnetic fields will actually cancel each other out. As for the motors putting capacitors across them should help out on interference. As for the price I have found a company that makes fairly cheap compass sensors. Yeah this is not for first but another robotics project of mine.

[Dave Flowerday]

Originally posted by wysiswyg
As for the motors putting capacitors across them should help out on intefearnce.

It's not the interference so much as the magnetic field created by the motors. Unfortunately the magnetic field is what makes the motor turn so it's a little tricky to filter it out

[Adam Y.]

It's not the interference so much as the magnetic field created by the motors. Unfortunately the magnetic field is what makes the motor turn so it's a little tricky to filter it out

Hmmm lenz's law. Always hurts never helps. We need another magnetic field to cance out the motors. Ill have to check my books there must be something in there on how to get rid of interference of motors. The wires was the easy part. Ill get back to you after I read all my books. By the way, wrapping wires together is generally a good idea to reduce interference.

[Elgin Clock]

Look Here

We did something with a magnetic sensor last year (2002) for our customized I.D.A.N. (Intelligent Detection, Analysis and Navigation) system.

Read my post there and if you want more info please let me know and I will see what I can do.

[Al Skierkiewicz]

Originally posted by wysiswyg
The wires was the easy part...By the way, wrapping wires together is generally a good idea to reduce interference.

Wrapping the wire as such does little to help with the induced fields. You must either shield with a magnetic and electrostatic shield or run balanced lines to cure that problem. The rise time of the pulse is an issue. But that is not the dominant component. The field from the motor magnets themselves are orders of magnitude larger than the earth's magnetic field so they tend to polarize the compass. If you could shield the motors and build the robot out of non-ferrous metals you might have a shot but you still run up against the variables of the building construction getting in the way of the earth's field. Although it hasn't been mentioned here yet, did you know that the earth's magnetic pole is not the north pole? If you are trying to base coordinates on an earth reference you must make corrections depending on your longitude. In Illinois we are almost directly south of the mag pole so no correction is needed but in the NW US or Alaska the mag pole is due east or maybe even southeast.
We did come across several compass chips that had reset lines to pulse the chip and surrounding area to kill the local fields and try to take a reading. That pulse was in the two amp range and required some external circuitry that was expensive and weighty. After weeks of walking down the compass path, we abandoned the research in lieu of a far simpler approach, the gyro. Yes it has it's faults but the accumulated errors over a 15 second period are small.
If you are still researching, check out magnetic shielding products. They are usually high nickel and laminated steel products. You may be able to put a shield between the robot and the compass. There must be a way to do it because my Dodge van has a compass that works fairly well although it only gives eight directions.