Do you suppose (assuming I could power them legally) that it would be legal to attach 2 old style mercury thermostats to our robot. Any rules that would prevent this?
Mercury is a hazardous material, and likely would not be allowed on the robot.
But is there a rule?
[R08] …
There really is only a tiny amount of mercury in them. About the same as a normal thermometer. Also, did you know that there is mercury in every flu shot… How about circuit boards. There’s mercury in some of them. There are also radio active isotopes in many modern circuits. I guess I want to know how much is too much if there is a rule.
Personally, I would say that it is not intended to apply to this. I suppose that I’ll have to ask Q&A.
Ask the GDC if you want an official ruling. Look up R08, and by the way, there is a lot more mercury in the thermostat than a thermometer. :] 
The other question to this is what are you trying to acomplish by attaching them? Some members of the CD community may be able to help you come up with a better solution. Give us more detail about what you want to do with them and let us see if that is the only solution. My guess is that the GDC will tell you no, but I have been wrong before.
Sounds like a tilt sensor. Not only is mercury a haz mat but, it would slosh around and not work well. A highly filtered accelerometer would work as a tilt sensor.
Wil,
Mercury is not allowed under the referenced rule. Among other things, it is impossible to get out of the playing field carpet. As Gary has pointed out, there are several alternatives for what your intended use might be.
On another channel, a “digital gyro from this year’s kit of parts” was suggested. Is that also equivalent? And how hard will it be to make work?
Why not use a ball bearing in a tube instead?
The kind of gyro supplied as an FRC kit part will not work well as a tilt sensor. You definitely want an accelerometer.
You don’t happen to know what it reports do you?
What if you mounted it on it’s side?
The KOP “gyro” is actually a yaw rate sensor. Its output signal is a voltage which changes based on how quickly it is turning around its axis of sensitivity. The cRIO FPGA firmware has support for integrating that signal to compute a heading angle, but that angle is subject to a couple of errors. One error source is slight drift in the signal offset from the sensor. Over the fifteen seconds of autonomous robot operation at the beginning of the match, that error is usually not significant, but it can drift noticeably during a two-minute match. The other source of error is sudden brief shifts in robot orientation that can exceed the maximum range of the sensor. The computed angle will start losing accuracy when that happens. Again, it’s usually not a big deal for the first part of a match, but it can add up to a problem in a couple of minutes of constant motion.
If you mount it on its side, it will still tell you only how quickly it is changing orientation. It does not automatically provide an absolute indication of its angle. A two-axis accelerometer, on the other hand, can do a very good job of identifying the “down” direction.
Thanks Alan,
Extremely helpful! As all of you have been.
Neal