We had and arrangement that kinda-sorta worked for mid positions. I'll try and describe it. (This is from memory, and could be slightly wrong.)
Take one solenoid and plug an exhaust port. Connect the output of this solenoid that uses the plugged exhaust port to vent to the input of another solenoid. The other output of the first solenoid needs to be plugged to prevent venting the supply line into space. The outputs of the second solenoid go to the piston. To extend the piston, set the second solenoid to the direction that will extend the piston, and set the first solenoid to let air into the second. To retract, do the same but with the second solenoid in the other position. To stop movement, set the first solenoid so that the air supply is output to the plug and the exhaust from the second solenoid and therefore the piston is exhausted to a plug, which should trap air in one side of the piston. It may be beneficial to flip the second solenoid when you stop to keep air in the side of the piston that pushes outward trapped instead of the side that pushes back. Sorry for this being so confusing, but I can't think of a better way to express my crazed thoughts.
I'll try and draw a diagram (let's see if my ASCII-fu is with me today):
Code:
--------
| |output plug
vent| |--| --------
| |--| | |
-------| / | vent| |---------
supply | | | |to piston
-------| \ |--------| / |---------
| |out - in| |
exht|--| |--------| \ |---------
plug|--| | | |to piston
-------- vent| |---------
solenoid | |
--------
solenoid
Again, I'm not quite sure if this is exactly what we did, some of the plugs may have to be in different places. I can't remember the exact workings of the solenoids right now. It was never a good idea to start with. I feel it needs repeating (since we missed this obvious fact last year) that air will compress, making the piston movable even when stopped with this type of control. The movement from my harebrained scheme is jerky and hard to use. Lastly, completely overshadowed by the other problems, any leaks will also affect the piston more as it will not be recharged by the compressor / supply until it is moved again.
All that said, it
is a way to achieve mid positions with FIRST legal parts. One last point is that it will help to apply a force to the piston so that there will be something to keep pressure in the piston when it is in midstroke, such as a bungie cord attached to the piston's end that will constantly pull it back (an idea one of our mentors came up with to save my unworkable midstroke arm controls).
And lastly, I have to say that I would never recommend this setup as a viable solution. It is way too much of an ugly, inherently uncontrollable hack. The arm based upon this idea never worked well enough to use effectively (and was removed before competition due to weight problems.)