Shooter consistency

[Mongai]

Originally Posted by Ether

If the spokes are not wide enough, that detector will have difficulty keeping up at high speeds.]

We had great luck putting a piece of reflective tape on our wheel and having the sensor count rotations of that instead.

[Ether]

Originally Posted by Mongai

We had great luck putting a piece of reflective tape on our wheel and having the sensor count rotations of that instead.

Good to hear.

Would you mind answering a few questions to make this useful for other teams?

1) What optical sensor are you using?

2) What angle of arc is subtended by the tape at the radius where the sensing is taking place? In other words, what is the width of the tape (in the circumferencial direction) where it passes under the sensor, and how from the center of rotation is the sensing element located?

3) What wheel speeds were you measuring?

4) Were you indeed "having the sensor count rotations" or were you using the FPGA to measure the period? (That's an important distinction).

5) What speed control algorithm did you use?

[Mongai]

Originally Posted by Ether

Good to hear.

Would you mind answering a few questions to make this useful for other teams?

1) What optical sensor are you using?

2) What angle of arc is subtended by the tape at the radius where the sensing is taking place? In other words, what is the width of the tape (in the circumferencial direction) where it passes under the sensor, and how from the center of rotation is the sensing element located?

3) What wheel speeds were you measuring?

4) Were you indeed "having the sensor count rotations" or were you using the FPGA to measure the period? (That's an important distinction).

5) What speed control algorithm did you use?

I can definitely get an answer by tomorrow. I wasn't the one who worked with the optical sensor.

[F22Rapture]

Would voltage control mode on a CAN Jaguar be consistent enough to get away without a feedback loop? The way our shooter was done, I can't really see an easy way to add one. Transmission used belts, and the wheels we were given are Aluminum (so no magnetic sensor), and an encoder mount is impossible without redesigning the entire gearbox. We also used pneumatic tires with a custom machined rim (so no holes to track with an optical sensor). Unless we attached a magnet to the wheel and used a hall effect sensor below it... but that would destabilize the already fairly shaky wheel unless we used a *tiny* magnet.

[Kevin Sevcik]

Originally Posted by F22Rapture

Would voltage control mode on a CAN Jaguar be consistent enough to get away without a feedback loop? The way our shooter was done, I can't really see an easy way to add one one. Transmission used belts, and the wheels we were given are Aluminum (so no magnetic sensor), and an encoder mount is impossible without redesigning the entire gearbox.

Betcha you could put a piece of reflective tape on your belt and use a IR proximity sensor on that. Don't get too attached on putting the sensor on some rotating thing, yeah?

As for the voltage control on a CAN Jaguar, that's better than nothing, not as good as actual feedback control. It will compensate somewhat for state of charge differences in your battery. It won't compensate for the spin down from firing a frisbee and it's going to deal poorly with rapid voltage changes from, say, your drivetrain or some other system making voltage fluctuate. Feedback is really the best way to go for getting fast, consistent shooting.

[Ether]

Originally Posted by F22Rapture

Would voltage control mode on a CAN Jaguar be consistent enough to get away without a feedback loop? The way our shooter was done, I can't really see an easy way to add one. Transmission used belts, and the wheels we were given are Aluminum (so no magnetic sensor), and an encoder mount is impossible without redesigning the entire gearbox. We also used pneumatic tires with a custom machined rim (so no holes to track with an optical sensor). Unless we attached a magnet to the wheel and used a hall effect sensor below it... but that would destabilize the already fairly shaky wheel unless we used a *tiny* magnet.

Paint the wheel black and then attach a piece of reflective tape and use an optical sensor.

[F22Rapture]

Would the line tracking photoswitch from 2011 work, or is there another sensor that would be a better choice?

[Kevin Sevcik]

Originally Posted by F22Rapture

Would the line tracking photoswitch from 2011 work, or is there another sensor that would be a better choice?

Yes, though it's a little slow. You typically need a pretty wide stripe to get it to work. Too narrow a stripe and it won't register. Practically speaking, you can make the stripe take up half of whatever surface you're sensing and it'll work fine.

[DampRobot]

Be fair warned, our software folk had a ton of trouble with the battery voltage dropping too low when we used the KOP photogates, to then point where they gave bad data. We were running 2 CIMs on our shooter, and the voltage the sensors were seeing dropped from 13v to 10v when we spun them up to full speed.

[Ether]

Originally Posted by DampRobot

our software folk had a ton of trouble with the battery voltage dropping too low when we used the KOP photogates,

Read the last two paragraphs on page 2

Not sure if this rule exception is valid in 2013.

[fb39ca4]

Originally Posted by Ether

Read the last two paragraphs on page 2

Not sure if this rule exception is valid in 2013.

I don't see anything in the rules about it, so I would say no. It wouldn't hurt to contact FIRST and see if they can add the exception back in.

[Ether]

Originally Posted by Kevin Sevcik

Yes, though it's a little slow. You typically need a pretty wide stripe to get it to work. Too narrow a stripe and it won't register. Practically speaking, you can make the stripe take up half of whatever surface you're sensing and it'll work fine.

FWIW, the Rockwell catalog for the 42EF model sensor lists the "Response Time" as "1 to 16 ms".

It's not clear whether those numbers mean

a) the pulse width must be 1 to 16ms in order to be reliably detected, or

b) it takes the sensor 1 to 16ms to respond to a pulse (and the required pulse width is left unspecified, or

c) something else.

You need the stripe width to be 1/12 of a revolution at the point of sensing in order for it to be under the sensor for 1ms when the wheel is spinning at 5000 rpm.

[faust1706]

many teams had great accuracy this year. What was really amazing is that both main types of shooters, linear and semicircle, worked great. 1706's shooter was deadly this year, but sadly teams caught on in eliminations and played excellent defense, a well earned victory. http://www.youtube.com/watch?v=U5PgpTQCgBc

[MUDavid]

This is what we learned at our 2nd MAR district and we ended up with the most Autonomous points. We had the same problem. Make 3 in Autonomous and then for some strange reason, frisbees were high. So we said, slow down the speed. Didn't seem to help. Our analysis was that at slower speeds with the frisbee encountering a lot of air resistance, the frisbee had a tendency to "float" and end up high. We cranked up the speed (within reason - there may be a point of diminishing returns) and adjusted our angle. If you can not adjust your angle, then position your robot differently (works great when shooting from the side of the pyramid). This works. Watch how other teams shoot. Keep it simple and good luck!

[dheerm]

our team was having trouble with encoders as well. Yes different battery voltages do matter when sending a value to a Victor because the value sent to the victor is the equivalent of a percentage of your current voltage. So to try to get a better measurement, I tried a last minute fix.


Wheel Motor Voltage = Value sent to Victor * Battery Voltage.

Therefore, Value Sent to Victor(Or other speed controller) = (End Voltage / Battery Voltage).

All you have to tune is the End Voltage that you want your wheel to have. Batter Voltage can be calculated (in Java) by instantiating a DriverStation object and using the getVoltage() method.

Same method should be in other API's.

It didn't give us 100% consistency but it did significantly improve it. There are of course other variables.