VEXpro 2016

[nuclearnerd]

There's a big asterisk on the "absolute encoder" thing. It's still a relative encoder, but it includes it's own counter. It's nice not to have to implement a counter on the rio, but you would still have the problem of the zero changing every time you start up the robot. An actual absolute encoder would give the proper reading no matter how often the sensor had been turned off and moved. (see page 22 of the encoder user guide here: http://content.vexrobotics.com/vexpr...User-Guide.pdf)

I'm wondering - do the rules allow us to power the encoder from a separate 5V battery that we could leave on all the time (even when the main battery is removed)?

[ahartnet]

Originally Posted by nuclearnerd

There's a big asterisk on the "absolute encoder" thing. It's still a relative encoder, but it includes it's own counter. It's nice not to have to implement a counter on the rio, but you would still have the problem of the zero changing every time you start up the robot.

Thanks for the catch. Definitely a big asterisk that I'm glad you pointed out. Got wrapped up in too much excitement!

[nuclearnerd]

On the subject of the encoder - it looks like it needs a "diametrically polarized" magnet to spin in front of it to work. The user guide I linked has lots of tips for embedding the magnet in existing gearboxes, with the glaring exception of VEX ballshifters (which currently need encoders to have a 1/4" bushing-supported shaft).

I wonder if VEX has a solution already cooked up for ballshifters and these new encoders. If not, I wonder if it would be possible to make a replacement for the plastic encoder gear that has pockets for multiple magnets around the perimeter (or something).

[Munchskull]

Originally Posted by nuclearnerd

On the subject of the encoder - it looks like it needs a "diametrically polarized" magnet to spin in front of it to work. The user guide I linked has lots of tips for embedding the magnet in existing gearboxes, with the glaring exception of VEX ballshifters (which currently need encoders to have a 1/4" bushing-supported shaft).

I wonder if VEX has a solution already cooked up for ballshifters and these new encoders. If not, I wonder if it would be possible to make a replacement for the plastic encoder gear that has pockets for multiple magnets around the perimeter (or something).

It has one.

Edit: Ops thought you were talking about the VP encoder stage.

[nuclearnerd]

Originally Posted by Munchskull

Edit: Ops thought you were talking about the VP encoder stage.

Nope, the ball shifters. Although with a little creativity, maybe the versa planetary stage could be hacked onto the ball shifter...

[Thad House]

One thing I'm wondering now is why the 775 didn't get released last year? Its part number is right in the middle of last years numbering, and its drawing says October of 2014.

[sanddrag]

It'll be fun to swap a couple of these new 775s into our 2015 stacking robot, which currently uses two BaneBots 775s on the lift. Love the specs on this new motor. In theory, you could do a full robot without CIMs, though I wouldn't recommend it. I'm interested in seeing how this plastic tubing holds up too.

[T^2]

Originally Posted by Thad House

One thing I'm wondering now is why the 775 didn't get released last year? Its part number is right in the middle of last years numbering, and its drawing says October of 2014.

Probably took a while to get approval from FIRST. 1323 was testing them on their bot during Madtown Throwdown 2014.

[ozrien]

Originally Posted by nuclearnerd

There's a big asterisk on the "absolute encoder" thing. It's still a relative encoder, but it includes it's own counter. It's nice not to have to implement a counter on the rio, but you would still have the problem of the zero changing every time you start up the robot. An actual absolute encoder would give the proper reading no matter how often the sensor had been turned off and moved. (see page 22 of the encoder user guide here: http://content.vexrobotics.com/vexpr...User-Guide.pdf)

The sensor provides both an absolute and a relative position. The "absolute position" is an absolute 12bit encoded value that wraps every rotation. This works similarly to the MA3 absolute analog encoder which gives you an absolute voltage (0 to 3.3V) for one full rotation.

You can think of it like having an MA3 analog encoder and a relative quadrature encoder both kitted in one, which solves several challenges in closed-looping.

If you don't move the mechanism and you power cycle the sensor, the bottom 12bits of the analog position remains unchanged.

Yes both the absolute and relative positions count the wrap arounds. But this doesn't mean you lose the benefit of absolute position. If you're application does not need wrap-around counting (like a continuous swerve) its pretty easy to add in the number of whole rotations to your 12 bit target position to achieve what you want. In fact this gives you the benefit of absolute position, plus you get to decide which direction you want to servo (always clockwise, always counterclockwise, or choose the path with smallest distance.)