Re: Pneumatic speed
 

Not over-pressurizing it, but giving it time to reach full pressure.

For people who don't understand this: A pneumatic cylinder will normally start moving as soon as air starts to flow into it, If you lock a pneumatic cylinder in a slightly extended position and allow it to reach working psi, it will throw with a lot more force than just turning on air to it and having it extend slowly as air flows into it.

Re: Pneumatic speed
 

This is correct. The physics don't allow my 60psi working air pressure to magically become higher when valve is opened and movement of the piston is mechanically restricted. It simply allows for a more instantaneous punching action

Re: Pneumatic speed
 

Actually the piston is receiving 60psi the instant the solenoid switches. The reason it shoots slower in normal function is because the air on the other side of the cylinder has to be pushed out by the new air coming in. It takes time to do this and force. If you stall the piston's extension in the manner you're talking about then this allows for the air on the other side of the cylinder to be released. Pressure does build up on the other side, but all the air technically doesn't go into the piston. After the shaft is released does the remaining air enter the cylinder.

Re: Pneumatic speed
 

Another option that 225 has been experimenting with is using exhaust valves on the cylinders themselves. This allows the pressure to vent directly out of the cylinder, and not have to flow back through the solenoid.

Re: Pneumatic speed
 

Sounds like 3138's 2010 kicker, which could kick a soccer ball cross-field with ease.

Re: Pneumatic speed
 

no, the cylinder receives only 'flow' from a 60 psi reservior. the pressure in the cylinder will only be what is required to move the piston forward until the piston becomes either obstructed or fully extended. This pressure may be as low as one psi delta.

What I'm saying is that if you just hold the piston in place for a short bit, the pressure will be able to build up. The business about the air getting out of the back side of the cylinder is so not-important. If you build-up 60 psi on one side and have effectively zero on the other side, it will move very fast. Our prototype can shoot across the room. But accuracy and repeatibility are still in question.

Re: Pneumatic speed
 

You're talking about building up pressure in the lines to the air cylinder, essentially. There is little volume in those lines (or, at least there shouldn't be!) and thus it won't help the cylinder deploy a whole lot faster.

The 'right' way to solve this problem IMO is to increase flow to your mechanism (better valves, shorter tube runs, multiple storage tank taps, multiple norgren regulators, multiple smaller cylinders to utilize more valves) rather than adding a whole other mechanism to restrain a cylinder.

Re: Pneumatic speed
 

You can actually build up pressure in the cylinder itself if it isn't contracted all of the way when you start filling. In 2008 lots of teams did this to allow the cylinders to expand faster. I posted a picture of one such catapult in this post.

Re: Pneumatic speed
 

Well, clearly it works, so there's not arguing against success. It still doesn't seem like a great solution to me, I would prefer to have single-step mechanism actuation for reliability reasons.

Here is a post with 95's 2008 shooter. Just lots of pistons and valves!

Re: Pneumatic speed
 

The reason the cylinder is slow is because it is pushing a lot of air through the vent, the fittings, the tubing, and finally the solenoid block (the same is true for the air flowing into the cylinder). So you can increase the psi and the size of the solenoid, but this will always be the main limiting factor.

The solution is to precharge the cylinder before firing it, and only plumb it in the firing direction (leaving the vent wide open). This method requires a locking mechanism (like a gate latch) to hold the cylinder back when precharged, and something to retract the cylinder (like a smaller diameter cylinder that gets out of the way). You can even directly plumb an air tank to the cylinder to decrease the psi drop during extension.

I'm sure you could improve upon this idea, but it's at least a starting point.

Re: Pneumatic speed
 

Marc S. was very close to the idea that my team 1108 used to win the Crossroads regional.

If you only retract the piston partially, you leave a reservior right in the cylinder to fill with air. Fill that to 60 psi while you hold down the catapult (we use an electromagnet) and then let go.

Also, yes, you don't want anything at all on the outputs of the cylinders. We simply use gravity to retract.

Of special note, the two robots that finished Crossroads as top seeds both ended up in the finals were both fully pneumatic.

Our 1108 has the system described above, where 1024, Kil-A-Bytes, has a circle of six very small diameter cylinders with very big valves and they shoot very well too, without a failure as far as I know.

I'd be curious to know how many other robots out there are full-pneumatic and how they've done.