AndyMark RAW Box and Potentiometer

[vigneshv]

Hi!
We're using an AndyMark RAW Box as the gearbox for our angle changer mechanism for our shooter, but we want to have a reliable way of determining the angle of the shooter, which will not depend on position at the start of the match (so preferably not an encoder, we're leaning towards using a potentiometer). Have any teams successfully attached a potentiometer to an AndyMark RAW Box, and how have you done it (or any gearbox with a threaded output)?

[Knufire]

Since the AM RAW box already has the encoder mount for the E4T, you might be able to add a limit switch or hall effect sensor somewhere else on the mechanism to give the encoder a reference point. This way, you can use the encoder for absolute position. We used this approach on our 2015 elevator and it worked perfectly.

Another option is to use the AM String Potentiometer and attach it to the base of your robot and the string to the mechanism: http://www.andymark.com/product-p/am-2674built.htm

[vigneshv]

Originally Posted by Knufire

Since the AM RAW box already has the encoder mount for the E4T, you might be able to add a limit switch or hall effect sensor somewhere else on the mechanism to give the encoder a reference point. This way, you can use the encoder for absolute position. We used this approach on our 2015 elevator and it worked perfectly.

Another option is to use the AM String Potentiometer and attach it to the base of your robot and the string to the mechanism: http://www.andymark.com/product-p/am-2674built.htm

So would this be implemented by moving the mechanism towards the switch at the beginning of the match, and then resetting the position of the encoder to 0 when it reaches the switch (because as far as I understand the only information you can get from both limit switches and hall effect sensors is on/off)?

[Knufire]

Originally Posted by vigneshv

So would this be implemented by moving the mechanism towards the switch at the beginning of the match, and then resetting the position of the encoder to 0 when it reaches the switch (because as far as I understand the only information you can get from both limit switches and hall effect sensors is on/off)?

Exactly!

[vigneshv]

Originally Posted by Knufire

Exactly!

Thanks, that sounds like it'll work!

[tr6scott]

Watch speed and backlash.... That homing routine is ok for most FRC applications, in industrial servo systems, the homing routine is...

Advance toward switch high speed.
Trip switch,
reverse direction,
run slow,
release switch.
set zero.

This method puts the gear backlash behind you...
The direction you need to move off zero, the gears have already meshed to drive that direction.

Probably over kill if you only have the one gearset, but if you have multiple gears, the backlash adds up.

[Kevin Sevcik]

Originally Posted by tr6scott

Watch speed and backlash.... That homing routine is ok for most FRC applications, in industrial servo systems, the homing routine is...

Advance toward switch high speed.
Trip switch,
reverse direction,
run slow,
release switch.
set zero.

This method puts the gear backlash behind you...
The direction you need to move off zero, the gears have already meshed to drive that direction.

Probably over kill if you only have the one gearset, but if you have multiple gears, the backlash adds up.

You forgot looking for the index pulse on the encoder before setting zero.

Also, you might try adapting the e4p mounting for an SRX Mag Encoder You can use that in absolute mode. The mounting should be quite similar to the e4p, but you'll have to drill a hole in the shaft to put the magnet in.

[jvriezen]

Originally Posted by vigneshv

Hi!
We're using an AndyMark RAW Box as the gearbox for our angle changer mechanism for our shooter, but we want to have a reliable way of determining the angle of the shooter, which will not depend on position at the start of the match (so preferably not an encoder, we're leaning towards using a potentiometer). Have any teams successfully attached a potentiometer to an AndyMark RAW Box, and how have you done it (or any gearbox with a threaded output)?

You mileage may vary, but we had a raw box for our shooter arm angle (it also had the ball pickup on the same arm) and we destroyed a gear in the box at a week zero event. We've replaced the gearbox with a different solution and also lightened the arm somewhat.

[vigneshv]

Originally Posted by jvriezen

You mileage may vary, but we had a raw box for our shooter arm angle (it also had the ball pickup on the same arm) and we destroyed a gear in the box at a week zero event. We've replaced the gearbox with a different solution and also lightened the arm somewhat.

Hmm, we haven't run into that problem yet, but what's the alternative you are using?

[nickbrickmaster]

Originally Posted by vigneshv

Hmm, we haven't run into that problem yet, but what's the alternative you are using?

We switched to using a Toughbox Mini and running the motor parallel to the shaft instead of perpendicular, which gives more room for electronics as an added bonus. We've been having problems with sag, but we are using the encoder-and-limit-switch method described earlier in the thread, so we can use PID stabilization. It's also a slightly lower reduction, but still powerful enough, in conjunction with a 25:1 VersaPlanetary.

Our arm assembly was also very heavy at that point because it is an intake and shooter. We lightened the intake considerably after that event. Depending on the weight of your arm/violence of movement, a raw box may be sufficient.

[BeardyMentor]

Originally Posted by Kevin Sevcik

You forgot looking for the index pulse on the encoder before setting zero.

Also, you might try adapting the e4p mounting for an SRX Mag Encoder You can use that in absolute mode. The mounting should be quite similar to the e4p, but you'll have to drill a hole in the shaft to put the magnet in.

I will second the use of the SRX Mag Encoder. If you have access to a lathe, it is very easy to install and plugs right into the talon srx which frees up pins on the RoboRIO. a+ product so far.