controling extension on piston

[m1k3]

how do you control how much a piston extends???

[etphonehome]

you just custom order the piston as however long you want it to be

if you're trying to use an old piston that isn't quite the right lenth, but a little too long, put a mechanical stop in front of the piston so the piston stops
make sure you dial the flow controls way down so you don't wreck the stop or your piston when it hits

if the piston is too short just screw something onto the end to make it longer

[ay2b]

how do you control how much a piston extends???

In short, you can't. The pistons go all the way in, or all the way out, until hitting a mechanical stop (such as the end of the stroke of the piston).

There are complex things you can do where you get the timing just right, and make it stop in the middle, but this is not what they are designed for, and is not recommended. I only mention this because if I just said "you can't", some one would mention this as a way that you can.

[Racer26]

Take a look at 1075's 2007 arm. If you can find any photos from Greater Toronto Regional, from Toronto, Canada, or Kettering Kickoff offseason event, from Flint, MI, OR an event that happened in october/november in New Jersey (sorry, I dont remember its name)

We used an elaborate system of valves to gain much control over our airflow. We controlled the feed side, as well as the exhaust side, and used some fancy code to switch them quickly in such a fashion that avoided any 'drifting' phenomena as well as any oscillations. We had just a few buttons that would send it to preset positions. 'Top' for the top spider on the rack 'Middle', 'Bottom', etc. The judges thought it was cool enough to give us the Rockwell Automation Innovation in Control award. Of course, the system came with its share of problems too. We had a manual override joystick to override the presets if we so desired, however, if the trim got nudged, it caused the arm to act extremely unpredictably, and until we learned what caused the problem, it was difficult to diagnose. Also, the potentiometer we used for position feedback was prone to... problems.

We used a similar prototype in the 2004 offseason after discovering our worm gear driven arm was hopeless, we replaced with a primitive version of our 07 arm, and it drove us to the win at the 2004 Canada's Wonderland Invitational.

I'm sure you can expect to see another pneumatic multi-position arm out of us in the future, maybe not for 2008 though.

[Nate Laverdure]

We used an elaborate system of valves to gain much control over our airflow. We controlled the feed side, as well as the exhaust side, and used some fancy code to switch them quickly in such a fashion that avoided any 'drifting' phenomena as well as any oscillations.
Team 612 made a multi-position pneumatic arm in the off-season between 2005 and 2006 just to see if they could do it. They designed the valve system by referencing ChiefDelphi quite heavily. I know that they controlled the exhaust side to stop the arm in various mid-positions, like this:

http://farm3.static.flickr.com/2366/2176920098_e1e5fcd398.jpg (http://www.flickr.com/photos/n8laverdure/2176920098/)

The arm was plagued by drifting and oscillations, as described above. Eliminating lots of excess tubing partly solved the "bounciness" problem.

[DonRotolo]

Of course, they also make double-solenoid valves that allow for mid-stroke stops, without the need for single and double solenoids controlling a single cylinder.

The truth, however, is that pneumatic pistons are really only good for all the way in and all the way out. It's possible to do mid-stroke positioning, but it doesn't work well, especially if the load on the piston is large (more than 5% of rated force) or variable.

Don

[Racer26]

Yes they DO make a double solenoid valve in which energizing both sides of the coil moves the setup to a locked position, but I think we came to the conclusion that they were either far too expensive to be practical for our team (we're pretty low budget, $10,000 or less on average per year, including our registration), or that they were not FIRST-legal, and as such designed our system using FIRST parts.

The primitive system we used in 04 was exactly what you drew out. Our 07 system was much more advanced with the valving, and the software much more sophisticated. Not only were we using the valving to control it, but we had coded it with the position sensor such that any outside forces (other robots, game pieces, team members, quantum particles) causing a change in the position of the arm would induce the arm to automatically correct itself. We thought this to be important when trying to place a tube on the rack, that the robot needed to maintain position with the arm. It worked fairly well, and was among the fastest arms we saw at GTR.

[m1k3]

Thanks alot for all of your responses i will attempt to follow that digram shown up there btw the single solonoid is hooked up to a seprate spike then the double solonoid correct?

[Jonathan Norris]

Of course, they also make double-solenoid valves that allow for mid-stroke stops, without the need for single and double solenoids controlling a single cylinder.

The truth, however, is that pneumatic pistons are really only good for all the way in and all the way out. It's possible to do mid-stroke positioning, but it doesn't work well, especially if the load on the piston is large (more than 5% of rated force) or variable.

Don

Don't dismiss multi positioning pneumatics system so quickly. After a unsuccessful arm system in '05 we rigged up our arm design with a multi positional pneumatics system and it worked surprisingly well with heavy use and only two tanks on board the compressor was strong enough to keep up with the load. For a game like this where you will only need two main positions for an arm system, a multi-positional pneumatics system for your arm is a very good option and far easier then implementing a beefy motor to handle the load. Of course there are negatives, the big one being bouncyness, but with a little creativity these problems can be solved.

[Nate Laverdure]

Thanks alot for all of your responses i will attempt to follow that digram shown up there btw the single solonoid is hooked up to a seprate spike then the double solonoid correct?
Correct.

[TubaMorg]

Don't dismiss multi positioning pneumatics system so quickly. After a unsuccessful arm system in '05 we rigged up our arm design with a multi positional pneumatics system and it worked surprisingly well with heavy use and only two tanks on board the compressor was strong enough to keep up with the load. For a game like this where you will only need two main positions for an arm system, a multi-positional pneumatics system for your arm is a very good option and far easier then implementing a beefy motor to handle the load. Of course there are negatives, the big one being bouncyness, but with a little creativity these problems can be solved.

We have also had good experiences using variable positioning pneumatics. We discovered last year that if we couple the pneumatic cylinder with a gas spring (i.e. like the ones that hold that hatch up in a hatch back car) we could get even more control and dampen any drifting and bounciness.

[m1k3]

anybody have some program codes for this?

[tseres]

Take a look at 1075's 2007 arm. If you can find any photos from Greater Toronto Regional, from Toronto, Canada, or Kettering Kickoff offseason event, from Flint, MI, OR an event that happened in october/november in New Jersey (sorry, I dont remember its name)

We used an elaborate system of valves to gain much control over our airflow. We controlled the feed side, as well as the exhaust side, and used some fancy code to switch them quickly in such a fashion that avoided any 'drifting' phenomena as well as any oscillations. We had just a few buttons that would send it to preset positions. 'Top' for the top spider on the rack 'Middle', 'Bottom', etc. The judges thought it was cool enough to give us the Rockwell Automation Innovation in Control award. Of course, the system came with its share of problems too. We had a manual override joystick to override the presets if we so desired, however, if the trim got nudged, it caused the arm to act extremely unpredictably, and until we learned what caused the problem, it was difficult to diagnose. Also, the potentiometer we used for position feedback was prone to... problems.

We used a similar prototype in the 2004 offseason after discovering our worm gear driven arm was hopeless, we replaced with a primitive version of our 07 arm, and it drove us to the win at the 2004 Canada's Wonderland Invitational.

I'm sure you can expect to see another pneumatic multi-position arm out of us in the future, maybe not for 2008 though.

were you the team with that really huge long piston? i just remember seeing one that looked like 3 feet tall......

(i was there and at waterloo, too, team 1565)

[MiniNerd24]

how do you control how much a piston extends???

If someone has said this before, let me know.
Try writing an algorithm that counts how long you press the activation button. Have it proportional to the air that will be "pumped" into the piston. Now you may need a component that limits air flow gradually. (sorry no idea where to get one.:o )

[Racer26]

were you the team with that really huge long piston? i just remember seeing one that looked like 3 feet tall......

(i was there and at waterloo, too, team 1565)

Our ram was about a 10" stroke. 1.5" bore. We had nice black carpet ramps at GTR... we took them off for Kettering and NewJersey because they were heavy and ineffective. We also added a braking mechanism to it in the offseason to further dampen any oscillations. basically a piece of aluminum angle passing through aluminum square tube with a ram to pin it.

[JesseK]

If you don't care about accuracy and this isn't for autonomous, you can try a Pilot Operated Check Valve. Came in the 2001 KOP, saw some on mcmaster a few moons ago.

[nickcvet89]

We had great success with using the optional magnet they give with the pistons, this is good if you have a piston that is a little to long for the job.