Arms

[meatmanek]

Originally Posted by Qbranch

Just as a rule of thumb... motors, because of the small compressor, tend to be faster at actuating manipulators n such, just because of the low cfm of the thompson compressor. We had a pneumatic actuator on our tetra bot... and it did slow us down some.

This year... motor drives on the elevator.... bottom to top (top row score) about 1.5 seconds give or take

I would have to disagree. Generally speaking, a pneumatic is faster than a motor. However, in some cases, you're correct. If you don't have any air tanks, you're limited by the airflow of the compressor. Also, if you're firing pneumatics quickly enough, or using a large enough cylinder that you're constantly draining your air tanks, the speed of the compressor comes in to play.

The trick is to have enough reservoirs, and to not use a larger cylinder than necessary.

Another nice feature about pneumatics is that they tend to be really fast when they're not under load, but they also can provide plenty of force. Our grabber can snap open, which is great when we're picking up tubes, but it also holds very strongly. It's difficult to have both with a motor.

And, generally speaking, you can get more power out of a pneumatic per ounce (excluding the compressor), which means a lighter robot, and more importantly, a lower center of gravity.

We used a combination of motors and pneumatics for our elevator this year. There's a minibike motor driving it most of the way, then when we hit the top of our 6' arm, we extend a 12" pneumatic extension, for a total height of roughly 8'. (For those of you who added 6'+12" = 7', we also have the length of the unextended cylinder in there)

[Andrew Schreiber]

For the last couple of years 27 has used no pneumatics (which is not related to the fact that we cant spell the word which would make our boards bad). Id have to say that it depends on what you are doing. For a one shot thing that has a straight line motion (Such as lifting) Id say it would be easier to use pneumatics. But the compressor weighs a lot. We had no where near the weight to put it on. Nor did we need it. If you have to have that motion in a line there are many ways of simulating it with motors.

Also positioning does come into play, if you were going to use it to position your arm over the legs... good luck. If you were just going to use a small cylinder to open and close your grabber, that would be pretty handy. Actually a motor is a more difficult choice in that case.

Really, it comes down to if the benefits offsets the cost in terms of weight. Pneumatics can be very useful but at the same time weigh a lot. Also consider that perhaps you dont have to use the compressor, you can use the tanks and just top them off. The final thought is that it comes down to personal preference, in our situation it wasnt required and could be done easier or more efficiently with motors.

[LightWaves1636]

We're motors all the way but we don't do straight motor, we add gears and belts to make the extension and lifts more efficent. Plus it helps to keep everything within size regulations.

[redbarron]

In 2005 we had a robot that was completely pnuematics (except the drive train).We had 4 motions in our manipulator and they were all done with minimal size pistons. The thing I havent noticed anybody mentioning is ratios. Cable & pulley systems paired with a piston can be very effective and in terms of weight compared to a motor with chain beneficial choice. We had an extension, which was moved up and down to get the tetras off the automated loading stations, and then an arm that was fixed on the side of our robot that picked them up off of the extension which was done with a "hooking" action also done with a piston. But unfortunately we overlooked the fact that swinging on a tetra from the side of your robot has disadvantage: lining up on the side of a triangle goal in the center from a distance creates one heck of an optical illusion.

I am not for or against either motors or pnuematics because like its been said you have to look at the advantages and disadvantages toward what you are trying to do in reference to what the capabilities are of each of these, but I just thought that I would point out the option of ratios to overcome the limit of how much extension you get from the piston.

[EricH]

Motors here. We've almost never used pneumatics for an arm. Fisher-Price motors are our motor of choice for lifting things, either by lift (put two on the same winch) or by arm (two on the same sprocket). The only things we use pneumatics for are shifting, the occaisional wedge/anti-tip device ('02-'04), and the occaisional arm extension. (This year, we have three cylinders. One is our grabber power and the other two shift.)

[Tom Line]

Two motors, globe + denso on on the arm joint, along with a gas shock to counterbalance the weight of the arm. Small motor + ball screw + transmission for the wrist.

A pneumatic cylinder extends and retracts the arm (only need it extended for pickup off the floor and for scoring on the top row) and pneumatic cylinder opens and closes the gripper.

Pneumatics are great for *FAST* response or for linear motion. However, for exact and quick, I'd always go with a geared down motor on a non-back-drivable ball screw or worm screw.

PID loops are great, but if you can avoid the complicated programming with a simple mechanical solution, and you have the weight, why bother?

[1359th Scalawag]

Our robot this year had no motors in the arm itself, (2 motors from the kit at the base of the arm) One motor would move the arm up and down and the other would pull a wire connected to an elaborate system of pullys that could make the arm telescope up to 3x it's origional length. As much as possible we usually try to not use pneumatics.

We made the hand as simple as possible by not putting any motors in it at all. It's just a peice of metal bent into a hook shape that would slide along the ground. It was actually very fast at both grabbing the tubes and putting them on the rack.

http://www.chiefdelphi.com/media/photos/28136

We also use no motors to deploy the ramp. The arm extends and catches on the ramp to push the ramp into position. We somehow came up exactly at 120.0lbs on inspection. If we had decided to add any more motors than we did, we wouldn't have passed.

[Frenchie]

Our arm this year uses a single joint to reach middle and low and an elevator to reach high. It is powered by two Keyang window motors.
The grabber itself is powered by a 3.5" stroke cylinder. General advices we could give from our experience with this season are:
1) try to keep the number of degrees of freedom to a minimum. The more joints you have, the harder it gets to control.
2) Grabbing should just be a matter of opening and closing the gripper. Getting over/under the ringer wastes a considerable amount of time (as we experienced with our initial gripper design, used at great lakes). The best grippers just have one side of the claw rest on the outer edge of the ringer and the other side of the claw close on the inner edge of the ringer (cf. 1114)

Francois.

[Billfred]

In four years of FRC, I've had a hand in three arms.

1293 in 2004: 1x1 box tubing mast (only up to the 5' mark), structure for grabbing the 2X ball from angle aluminum as I recall, homebrew linear bearings, winched up and down by a steel cable driven by a van door motor. Winched fine, just not effective at its task. Ditched Saturday at Palmetto.

1293 in 2005: 1293 gets up close and personal with the joys of 80/20. Three-stage lift, driven by timing belts and a van door motor. Worked beautifully, but hampered with drive issues (namely an inability to get it pointed the right way for most of the Palmetto Regional). The only real problem was that we'd frequently trip breakers on the van door motor, but only when stalled for a significant amount of time. Much of the lift design (and some components, as 80/20 linear bearings aren't particularly cheap) was recycled for their 2007 robot, Atlas.

1618 in 2007: Nobody on the team had ever done a jointed arm before, but it was the cheapest way we could figure out. A small CIM and a BaneBots handled the first stage of 1.25" PVC pipe, while the second stage was a Denso motor. Eventually, the second stage was turned into part of a gripper, and temporarily changed over to a Keyang motor before going back for mounting reasons. (There are plenty of pictures on CD-Media of the pre-gripper setup, and a couple of the later setup.) The arm never fully functioned in competition, but appears to be easily fixed once 1618 returns to work after their spring break. (Victors, lemme tell you...)

Of the three, my favorite remains 1293's arm in 2005--I'd reuse that design again in a heartbeat, resources permitting. But if you're on a budget, you can't beat the PVC arm.

[efoote868]

After '05, We decided to stay away from an articulated arm; we've had much better success with elevators. They (can) move much faster, are easier to control (two optical encoders and a nice PID loop).