Cylinder response time

Hi all,

maybe my question is not related directly to the PLCs but I have control problem. I am controlling a pneumatic cyclinder through an air actuated valve. The air piloting the cyclinder valves is switched ON/OFF by a Festo valve controlled by DeviceNET from an AB PLC. My problem is that the response time of the cylinder is quite low and not constant. My suspicion is that the air hose from the Festo valve to the cyclinder valve is too long.
Would it be possible that the length & diameter of the hose affects the response time?

If by reponse time how quick it extends or retracts then the simple answer is yes.

What you need to do is state size of cyclinder, air pressure setting, diameter of supply line(s), which valve and if possible what the load on the cylinder may be. Oh yeah are there any flow control devices in the system?

It could be as simple as turning a regulator up or using a larger hose. It could also be mis-designed and need a larger cylinder OR there may be an issue with DNET communication (this one I doubt).

Short answer yes. One solution would be to installed a volume booster at the valve.
Let the Festo valve control the volume booster. Booster would have its own supply

While the length and diameter of the line may have an impact on response time, for a specific length and diameter, the response time should be constant.

You should be looking at things like the quality of your air supply. Moisture in the line can make valves behave erratically. Is the supply volume sufficient for your needs? Is the pressure reasonably constant?

How much variation in response time are you seeing?

Reduce the trapped volume of air...

Move the valve to be as close as possible to the cylinder and shorten the hose to the minimum valve possible. By minimizing the trapped volume of air between the valve and the piston, you reduce the amonut of air that is takes to presurize the cylinder to the point where the cylinder will move. Having low friction pneumatic cylinders also helps. I always try to mount the valves directly on the cylinder See the link but be warned the picture is BIG.

ftp://ftp.deltacompsys.com/public/pneu/P5250009.JPG

Peter,
Is that a tempsonics linear encoder in the centre of the cylinder? If so has the Cylinder been machined to suit it?

danbente;

What is the Bore Diameter, Stroke, Rod size?

Is the cylinder mounted horizontal, vertical, other?

What is the load in both directions of travel. Weight?, Force to move?

Is the load constant or changing?

Is inlet pressure regulated to a setting less than the pressure line feeding it drops to during the time it is operating?

A general rule for air cylinder line size is to use the same size or larger than the cylinder's ports. i.e. If it has 1/2 inch NPT ports you will get best performance with 1/2 inch hose. This is especially true if the line is quite long. There is a formula for calculating this.
Two other things that you may want to check are:
1. Valve oriface size, if too small it will act as a flow restricter.
2. What is the volume of air available to the system?

Thanks all for your suggestions. Sorry for the lack of details.
The valve which is controlling the cylinder is a 5/2 valve with 3/8 NPT ports. In fact we replaced an old system (a TI PLC with a AB PLC)and we decided to change also some pneumatics (bad idea...).
Before, the cylinder was controlled by a 5/2 solenoid valve (spring return) triggered by a PLC digital output, the only thing we changed was that we put an air actuated valve (spring return also) instead the solenoid valve. The air for the signal comes through a 1/8 hose from the Festo.
Both the valve supply and signal come from the same source - we have 80 PSI. The length of the hose for the signal is about 40 feet, the same is the length from the supply to the valve. The distance between the valve and the cylinder is about 3 feet.
The cylinder is mounted vertical.
My main problem is not the response time (I can anicipate the command in the PLC) but the fact that the response is not constant.
I made some tests:
- for 1 minute, I triggered a 200 msec command every second -- sometimes the cylinder went at full stroke, sometimes it went half stroke, sometimes it didn't respond. If I decreased the command time at 150msec, most of the time the cylinder didn't move at all. If I increased the command time at 250msec, the cylinder responded always but the time it stayed at full stroke was visibly inconstant (I cant tell how much exactly but it was visible). Here I have to mention that the PLC cycle time is about 30msec and stays steady at this value all the time. I have no idea about the response time of the DNET module of the Festo block but it shouldn't be a problem (for an AB VFD I measured a max. of 30 msec delay on DNET so it should be OK, but maybe the Festo module is not so fast - I'll check the specs).

Thanks again all for your help.

Bud and Peter are the pros at Fluidpower so maybe they will get a better understanding of the system.

The response for extend and retract will depend on the volume of air that the valve can supply and how much the cylinder needs to obtain resonance. The hose(s) and their length will have a bearing on the supply capabilities.
NOTE: Reasonance is the highest frequency that a cylinder can retract and extend continously and constantly.

There are formulas for calculating the SCFM required for the cylinder. The SCFM will determine HP, valve used and hose(s) needed.

Bud asked about the specific factors needed to determine what is needed. I am thinking you may have a hose issue OR possibly the valves is not sized properly.

Can you connect the Festo DeviceNet output to a scope or a chart recorder to see if the solenoid is seeing the 200 mSec pulses?

Double check your DeviceNet wiring and terminations. It's possible that there is a high number of errors and retries leading to a long DeviceNet 'scan' (Time between successive updates of the same node). Do you have access to any DeviceNet diagnostic tools?

Are the Festo solenoids 24 VDC? Is your DC supply adequate to handle the load? If the voltage drops too much when you fire a solenoid, you can get a sluggish response.

danbente;

Your biggest problem is the 40 ft air pilot line. You could help the situation a lot by installing a Quick Exhaust valve at the Pilot Port of the 5/2 Air Piloted valve.

This would allow a fast exhaust of the pilot air and a quicker response for the Air Piloted valve.

The problem most don't realize is that air filling a line can be quite fast while exhausting that same air may start fast but as pressure drops flow slows and time to get to no pressure, or low enough pressure to allow the spring to return the valve to its start position, increases.

Regulate the Air Pilot Supply to a minimum that gives consistent shifting time will also be a help. Less air in less air to get rid of.

After all that I trust you don't hope to emulate the solenoid operatiion response capabilities because it will never happen. Air has its limits since it moves very slow in relation to electrics.

Try using 1/4" pilot lines and change your air valve to a double air piloted valve as the spool will shift with a lower force. The lower pressure will be there before the higher pressure required to overcome the spring.

Thanks again to all of you.

I'll try to use some of your suggestions. If I'll find the solution I'll post it.

Dan

hi Danbente,

I had the same problem on a cut-to-length machine that is fully operated by pneumatic cylinders, and what's Bud (fluid power) has suggested is the solution , before I apply a quik Exhaust valve to cylinders the machine wasn't working as it should, and we were facing inacurate movements, it was very slow , and the spped was frequently falactuating due to change in compresoor's pressure, even though you have a pressure regulator on the machine....output pressure varies, and a quik Exhaust valve has solved the problem.



hope this help.