chain mechanism to tilt panel

[zigerati]

We're about as rookie as a team can be and while we're making progress and have had some great help from other teams there are some areas we're still struggling with (even at this late stage!).

We have a set-up where we're trying to tilt a horizontal panel (32" x 23") through a range of about 90 degrees with 5 pre-set stages (front 1: fully tilted, front 2: partially tilted, flat, back 2: partially tilted, back 1: fully tilted). We have a chain underneath the panel connected at each far side that runs through a tensioner on each side and then to what we want in the middle to be a motor with a sprocket. Everything on the mechanical side is figured out and working smoothly but we have no idea what kind of motor and controller to use that can be limited to that range of motion in labview and how to attach the motor properly to the sprocket. We have quite a few different types of motors on hand from the original kit as well as First Choice including several types of window motors. The load is not too great as the panel is balanced well in the center. The chain is basically pulling the panel around a pivot point.

We could really use some pointers to first tell us if this is even possible with a motor/controller/sprocket and then secondly, if so, what is the optimal motor/controller pair to use.

Thanks in advance for any tips!

[railerobotics]

For your application I think a jaguar with an encoder reading values off of the rotating axle would be the best. The same result could be accomplished with a victor, or talon with the encoder wired to the Crio, but you would need to write the feed back loop while the jaguar has one built into it. (As far as I can remember there might be a simpler solution to an encoder feedback loop now). The motor is almost down to personal preference except that you will want it to be slow and controllable, as well as having enough force for your application. You can hardly go wrong with a bag motor on something like a 100:1 reduction. 160 rpm's and more than enough force to accomplish whatever you want as well as the same mounting holes as a cim.

Pictures are always a great addition as well, gives a better picture of what you are trying to convey

Edit: Links for bag motors http://www.vexrobotics.com/vexpro/mo.../217-3351.html

Gearboxes http://www.vexrobotics.com/vexpro/ge...planetary.html

Jaguars http://www.vexrobotics.com/vexpro/mo.../217-3367.html

One type of sprocket that could work http://www.vexrobotics.com/vexpro/sp...sprockets.html

[EricH]

Is it possible? Yes. You'll want an encoder or a potentiometer--probably the pot--connected to the chain somehow, and you'll possibly want a gearbox as well, which might make for a nice place to put the pot. (The other good spot might be on a tensioner, or a random sprocket in the middle of the chain.

As far as what is optimal, that would depend on what exactly you have available. Window motors are good at holding position, but other motors may also work.


Programming setup (basic): First, figure out what the sensor reads at each desired position. Store that somewhere in the code. Next, figure out the command trigger for each position. Store that as well. When a command trigger comes in, compare the desired position to the current position, and move the motor the right direction to match the current position to the desired position. Stop the motor right before it gets to the desired position (it takes a little bit to get the command through). I'm no programmer, so I don't know how to actually implement this in code, particularly in Labview.

[maths222]

For a rookie team, I would not suggest choosing a Jag because you can plug a sensor into it, as to use it, you would need to use CAN. That said, a Jag is not necessarily a bad choice, depending on what you have. (Unless you are using a window motor, as they seem to not like Jags that much.)

[railerobotics]

Originally Posted by maths222

For a rookie team, I would not suggest choosing a Jag because you can plug a sensor into it, as to use it, you would need to use CAN. That said, a Jag is not necessarily a bad choice, depending on what you have. (Unless you are using a window motor, as they seem to not like Jags that much.)

Ah yes I forgot that CAN was required. I am not a programmer so I will ask you...is can hard to implement in code?

[maths222]

My semi-informed understanding is that the code is quite similar with CAN or PWM; however, connecting the Jags to the cRio and configuring them can be a challenge.

[SenorZ]

We used a pot for our shooter last year. We ran the pot to analog input and had code read the value as an angle. The software (which I didn't write or ever see!) would power the motor (a PG-188 I think) forward or backward until the desired angle was read back.

We had settings for vertical, 45deg, 30deg, and I think a couple of others.
They used PID to control the angle and keep it constant because the shooter was not centered and balanced. If your panel is, it might not be hard to position it.

[maths222]

Did you 3d print those gears?!

[EricH]

Originally Posted by maths222

Did you 3d print those gears?!

If those AREN'T 3D printed, based on the picture, I want to meet the machinist. Actually, 3D printed gears/sprockets are great in low-load applications, like a sensor connection.

[SenorZ]

Originally Posted by maths222

Did you 3d print those gears?!

Yes, one of our mentors did with his own MakerBot. Worked great, never slipped... though there was no real pressure on them and they were working at low rpm.

[zigerati]

Wow - thank you all for the pointers. That is a huge help. The pictures from railerobotics are especially helpful. We'll start testing some of the ideas that were posted and report back. Thank you all for the help.