Possible to Adjust Speed

I was wondering if anybody knew of a relatively simple method of adjusting the speed or volume of air flow into the piston? Basically allowing the piston to shoot out real fast or to slowly shoot out.
Thanks for your help!

There are flow control fittings used for this purpose.

Any methods for changing the speed during a match?

From my unprofessional experience with pneumatics, there’s no way of adjusting air flow during a match. you can adjust it manually that mark was talking about, but thats a little hard to do when your robot’s on the field!

Think of a piston as a light switch… not a dimmer switch! A light switch, on/off

You can do this by using multiple Directional Controls (DC) and multiple Flow Controls (FC). To obtain three different speeds you could use two DCs and three FCs. Remember to only flow control the out flow of the cylinder and not the in flow. Connect the cylinder port to the “P” of DC1 (DC1-P) and FC1 to DC1-A. Connect DC1-B to DC2-P, then FC2 to DC2-A and FC3 to DC2-B. With this arrangement whenever DC1 is de-energized FC1’s setting will be used and therefore that speed. With DC1 energized and DC2 de-energized you will get the speed based on FC2’s flow setting and thus with both DC1 and DC2 energized you will obtain FC3’s setting.

This is fairly easy to program since the DCs are basic binary on and off devices.

If you need more speed settings just add more DCs and FCs in the same “logic”. You will have one less DC than FC’s. If you wanted say 6 different speeds you could do it with 5 DCs and 6 FCs.

ted yz

There are a few threads on this already, but we are thinking about using a servo attached to one (both?) of the flow control valves on a piston. Not ideal I know, but it won’t require many parts.

A servo won’t give your the range you would need. The flow control valves are needle valves that may require a few turns to get to the desired flow rate.

Looking at your problem from a different perspective, you might want to try two air feeds to the same cylinder port, since piston speed will be dependent on amount of air flow. Opening a second feed to the cylinder can increase your piston speed, as it may fill the cylinder more quickly. Limits are imposed by inlet/outlet port size and available pressure, and you should test the concepts by prototyping. Then you can manually adjust each feed to fine tune your speed.:slight_smile:

Al - This is true, but we only need a relatively small deviation in pressure. Thanks though!

Also, you COULD gear the servo up to get more spins out of it. Dunno if the servo would have the oomph to do it is all.

There are better ways to do it using multiple valves. 1075 has in 2004, and 2007 built multi-position pneumatic arms, which could be stopped in mid-travel. by pulsing a feed line, you can cause it to slow down

You could put a motor and encoder on the pressure valve then control the position of the motor. It’s almost trivial with the Jaguar using CAN. The biggest difficulty would be ensuring your start point is where you want it. You could run it to a stop of some sort to make sure you have the position correct.

Another option is to control the duration signal letting the air in. If it takes 250ms for a full stroke, then having an “on” signal that is shorter than that will reduce the power it has when it reaches the endpoint. This is probably the simplest method and requires the least parts, but it requires more fine-tuning to make it work as desired. It might be possible to even pulse width modultate the air intakes to achieve an overall slower speed. Perhaps you would have it forward for 20 ms and backwards for 10 ms and repeat that pattern until it reaches the desired position. This will use more air since you are using the air to oppose the movement of the pneumatic cylinder.


Try searching for “proportional pneumatic valve” on google. There are many different flavors. A common unit I have used for prototyping air actuated brakes on very large robots is a proportional flow control valve such as:

We have prototyped this in our system with general KOP pneumatics and it works well to control velocity of the cylinder.

One lesson learned: Make sure you double check the signal voltage before ordering parts.

Can Anybody please private message me and tell me how to post my own thread?

Is it legal to use more than one valve? This would allow twice the flow rate into the cylinder.
Would you have to make sure that the Cv was smaller?

The limiting factor is still the volume you can pass through the actuator port.

Is it legal? Yes.
Do you need to lower the Cv? No.

Will it produce twice the flow? no. This is due to a number of factors including turbidity (sp?) when recombining the flows, etc. Experiment with this, as it is a good learning point in fluid dynamics.

Hint: use ‘Y’ connectors :wink:

Yes, there is. Use a “Y” to connect your piston to the air source, one leg has a restrictor in it, the other does not. Use a check valve to prevent bleeding air from the “unactivated” side and there you go.

Feed air from the unrestricted side and you get a rapid deployment of the piston, and from the other side a restricted air flow reduces the speed.

This is quite easy to do.

  1. Y-tie your exhaust ports together on your piston control valve and run it into the pressure input port § of a second valve (call it valve2).
  2. On one output of valve2 (A or B), attach a flow control to meter the exhaust. Leave the other output port (B or A), and the exhaust ports open.

Now, when valve2 is set to the open position, the cylinder runs at full tilt (or at whatever rate the flow controls are set to on the cylinder itself, if included). When valve2 has the flow control switched in, your cylinder will “exhaust slowly”, based on the flow control’s setting.

Couple this with additional flow controls on either or both ends of the cylinder to adjust individual full speed ‘default’ flow rates in each direction (and considering the limited amount of total air we have anyway), and this should give you plenty of rate options. I’ve found it covers over 90% of typical FIRST multi-rate applications.

Note that many valves are “piloted”, which means that they need a bit of air pressure on the input § port to accomplish switching so YMMV with them. IOW, depending on what valve you use for “exhaust plugging”, if it is spring return or not, AND which way you default it, it may take a fraction of a second for the exhaust air to build up in front of it enough to switch the valve.

Therefore, with spring return valves I normally suggest the “unenergized” state to be the metered/restricted flow state, JUST to make sure it WILL build up enough air pressure to switch to the other state. Those valves will automatically return to the unenergized state (regardless of coil signal !) once the pressure drops enough. In this app, that is OK.

BTW, you COULD try to meter the source with a second valve, but if you try it and set the restricted flow rate too low, you will find under some pneumatics plumbing circumstances (and valve choices) the directional valve will fail to respond due to insufficient pilot pressure (or on spring return valves, it can go the WRONG WAY!). Switching a flow control in/out of the EXHAUST plumbing instead overcomes this problem.

Does this make sense?

  • Keith

What if you used a servo that will turn CW or CCW without stops? You could use a piece of tubing to match the servo output and the flow control knob.

Not sure of this year’s rules regarding servos. (CAN you use stronger ones THIS season, or continuous rotation servos? You couldn’t in previous years.)

Bear in mind that the kit regulators take a LOT torque to turn. Probably way too much for the basic kit servo.

Now you CAN easily make a servo in software with a pressure regulator, a window motor, a 10-turn potentiometer, and TWO pressure transducers (1 for output, 1 for SYSTEM pressure, to make sure you CAN reach it before trying).

In addition to a simple closed loop servo routine in software, you also need to do a few other things:
(a) put both software limits to the pot AND a mechanical friction/slip clutch between the motor and the pot/regulator set, to keep from breaking either one should the motor or software go crazy and attempt to turn it too far in either direction.
(b) monitor the full system pressure as well, to disable the servo routine when the system has insufficient pressure. Otherwise, you will “polarize” the servo wide open, trying to get it to a setting you simply can’t reach.

But there are simpler methods.

Example - Create a Buffered Supply Tank, at any programmable pressure.

Use a center-blocked valve (or a pair of kit valves), two flow regulators, a pressure transducer, one Clippard tank, and software. Software is similar to the compressor routine. In this case though, whenever it is out of range EITHER way, pulse one or the other valve coil to either fill or drain the tank until it reaches the desired pressure.

One flow regulator controls the tank FILL rate (NEARLY wide open), while the other controls the BLEED-OFF (probably a fairly low setting, to prevent TOO much air waste from a user’s “setting indecisions” :D). A software variable now determines the tank’s desired “set” pressure, and it “automagically tracks it”.

Simpler: Monitor the cylinder’s pressure with a transducer. When the cylinder reaches the desired force, just cut off its supply! (Again, use a center-plugged valve.)

FYI, a center plugged valve is one of the order options for valve parts in the familiar KoP SMC SY3000 series. See the SMC datasheets for more info.

Does this make sense?

  • Keith