Position control

B

Thread Starter

Bruce Cincotta

I have an application where I want to adjust the position of a leadscrew from 0-100%. I want to use an air motor in the application and am thinking of using an encoder to a SLC. A limit switch would indicate the 0% position and a number of turns would take me to 100%.

Has anyone ever tried this? what should I look out for?
 
> I have an application where I want to adjust the position of a leadscrew from 0-100%. I want to use an air motor in the application and am thinking of using an encoder to a SLC. A limit switch would indicate the 0% position and a number of turns would take me to 100%.
>
> Has anyone ever tried this? what should I look out for?


You don't mention the total number of turns you're expecting to make so I'll assume its a fair number.
Things to watch for:
- make sure total number of pulses in full stroke can be handled by the SLC counter card.
- don't exceed max count frequency of the counter card. It will drive you nuts because the system will seem to work sometimes but not others.
- if accurate positioning is needed, use two air solenoids piped in parallel. The larger one gives "fast" and the smaller one gives a "slow" speed as you approach the target position.
- if the budget allows it, an electronic proportional valve will let you also use a tachometer feedback loop to control leadscrew speed regardless of load or air pressure.
 
A

Anthony Kerstens

Don't immediately scale the encode value
into %. Instead, scale it into a number
of screw turns or length. This will help
in trouble shooting.

If you want % to make it simple for the
operators, then do that after.

Anthony Kerstens P.Eng.
 
B
Do air motors stop quickly? You might want to use a brake if you need more accuracy.

Bill Sturm
 
B

Brandon Ellis

Bruce:

This is an interesting concept. I have not tried it, but have often wondered what the result would be to tie an encoder to an air motor and
introduce a motion controller to control a proportional air valve. My line of thinking is that there may be a few problems. First of all, due to the compressibility of air, the accuracy may be less than reasonable. Secondly, quick stops from high speeds may be impossible therefore sending the motion controller into oscillation hell.

I realize that you are opting to use a SLC for the encoder input. While most motion controllers are designed to read up to an 8 Mhz (or more)
encoder input, I am not so sure about the limits of the SLC. This, of course, is based upon the speed of the encoder output pulses (i.e. Pulses per sec) which would be directly proportional to the speed you plan on spinning the air motor. In other words, how fast are you going to run the motor and what's the resolution of the encoder?

I am curious to see how it turns out, please let me know.

I hope this helps,

Brandon S. Ellis
Sales Manager
Robotic Control Group
700 S. Illinois Ave.
Suite A104
Oak Ridge, TN 37830
Tel: (865) 425-0301, Ext. 160
Fax: (865) 425-0268
http://www.roboticcontrol.com
 
D

David Leese Dresser Valve Div., Hallibur

I designed an electric actuator with a ball screw that is positioned with with a rotary encoder on the motor shaft. My customer designed the
feedback and control loop. The problem they have had is that on loss of power, they don't know position, and they sometimes drive full-speed
and full-force into the end stops, which can damage equipment. We are now trying to incorporate a linear transducer along the ball screw so that position is measured directly, rather than interpreted from motor revolutions.

The 0% limit switch will help you verify location, but depending on the application, you will need to be careful on your approach if you are unsure of position.

D. Leese
 
B

Bruce Cincotta

We would use an air motor running 1000rpm or so, reduced to 1-2 rpm through a gear reducer. We would incorporate a limit switch to limit the end travelat 0%. The actual output shaft would move very slowly.
 
K
Hello...

We have an application very similar to this. In our application we have a long leadscrew. This leadscrew adjusts the geometry of a crank arm fulcrum. We used to use a small stepper motor to turn the screw. The life of these steppers in a wash down environment was not acceptable. So I came up with the idea of an air motor (we used a nice little motor and gear box from Gast). We control the bi-directional motor with high speed solenoids from Mac mounted right on the inlet ports. With the solenoids right on the ports of the motor there is no line charge up or bleed down delays in the system. The PLC knows that for every Xms of solenoid on time we see X thou worth of travel in the leadscrew. When we power up the system the screw is sent full travel to a end of travel switch. The PLC then knows and controls the position of the screw from that point on. This meets our needs quite well and if the resolution required by your application is about the same as ours... then this may work for you as well.

Best Regards... Rick Kelly

Chief Technician
Natural Cuts
Cheese Operations
Kraft Canada
(613) 537-8069 V
(613) 537-8057 F
[email protected]
http://trondata.on.ca
 
ON loss of power the unit should stop. If power wqs lost to the controller or encoder the count and therefore the position is lost. Usually you will then go through the "Homing sequence that you started up with, which should prevent you from "slamming" into the stop... A limit switch is normally used to locate the homing point.
 
N
In addition to the problem below, I have used air on a number of occasions (usually at the clients request) for position control and have always found it sadly wanting. Problems include compressibility, and therefore non repeatability
of position and response, slow, lack of power at slow speed and unreliability due to air valves blocking partially or fully and causing intermitttant operation. I wouldn't use air again if i could help it.
neil j.
 
A

Anthony de la Rosa

sounds like what you need is a distance meter -- check out Sick Optics and Specra Precision. I use Spectra but for your application, an open loop
system like what Sick offers will probably do just as well

anthony
 
I agree that the PID loop tuning might be a problem because of the "sponginess" of the system response. But why not use a linear motor for the
job... There are a number of them out there including a cylindrical linear with plenty of torque and slow speed accuracy...

Bob Close
Precision MicroControl
 
Use an absolute rotary position sensor if you can gear it in. Otherwise I agree with the linear transducer. What's the stroke?
 
B

Brandon Ellis

Bruce:
Why are you leaning toward the use of the air motor? If this is just something you are trying then that's great, but if you are looking at
overall costs, the fact that you are having to use a 100:1 gear head (esp. if you use a planetary gear head) could turn this project into a costly one (really costly if it doesn't work like you had hoped).

You know your torque / speed requirements, but it sounds like that by using a servo motor (brush or brushless) you may be able to maintain the slow speed on the ballscrew and provide the required torque, all without the expense of such a large gear head.

Just food for thought,

Brandon S. Ellis
Sales Manager
Robotic Control Group
700 S. Illinois Ave.
Suite A104
Oak Ridge, TN 37830
Tel: (865) 425-0301, Ext. 160
Fax: (865) 425-0268
http://www.roboticcontrol.com
 
B

Bruce Cincotta

The biggest fear we have with electric motors is the environment. Generally, everything needs to be rated NEMA4 and the temp could be 100-200 degF.

> I agree that the PID loop tuning might be a problem because of the "sponginess" of the system response. But why not use a linear motor for the
> job... There are a number of them out there including a cylindrical linear with plenty of torque and slow speed accuracy...
 
D

Daniel Boudreault

Hello D. Leese,

Air is compressable. Not good for motion control. The leadscrew will take some of the forces off of the air motor, but not totally, and your controller will be oscillating the output a whole lot, if you are trying to maintain position...especially if your gain is high...

It just depends on what you are looking to do..so what are you trying to accomplish?

It would work if you don't need repeatability and accuracy.

Think of Air as a spring in a control loop. It will be impossible to control if your control loop is faster than your feedback loop. I wished I had my signal feedback textbook with me... I can show you the feedback loop formula, and show you that it will be unstable...that is if you need like faster than 1 second update on your positioning loop...

Dan Boudreault.
 
D

David Leese Dresser Valve Div., Hallibur

First, regarding the distance meter, our general goal is to have something that will install inside the actuator body tube that holds the ball screw and that is packed with grease. I don't think optics are a good choice in this case. We've discussed using a Balluff rod-style linear transducer with a magnet in the ball screw, but we are having a problem finding a place for the head of the sensor. We need a rod style sensor
with a housing that can mount remotely. Any ideas?

Second, regarding the comment in another message about the "homing sequence" to rediscover position after loss of power. As you design a clever homing sequence, consider what the
procedures will be after loss of power. I think most of the problems with our equipment occur when operators manually bypass the automation, which frequently occurs during and after loss of power. Operating manually has the effect of bypassing the homing sequence. The operators are often aware that they are operating blind, but they don't always appreciate the power in their hands--the speed and force of this particular ball screw--until they bottom out the ball screw
and damage something. More training and better procedures are one solution, and direct position feedback is another.

D. Leese
 
MTS makes several linear devices that work on Doppler pulse effect. There are various resolutions and they can be made to work with most controllers that accept encoder counts, although it takes some programming... There are
also LVDTs which are used in the aircraft industry.

Bob Close
Precision MicroControl Corp
760-930-0101
[email protected]
 
B

Brandon Ellis

>>The biggest fear we have with electric motors is the environment.
>>Generally, everything needs to be rated NEMA4 and the temp could be
>>100-200 degF.

Bruce:

There are some motor manufacturers out there who produce motors (perhaps not linear motors, but rotary for sure) that can handle that type of a high temperature environment. As far as the NEMA 4 rating, I assume that means the environment is wet. There are electric motors that can handle both of these aspects.

Kollmorgen certainly comes to mind, but I am sure there are more.

Brandon S. Ellis
Sales Manager
Robotic Control Group
700 S. Illinois Ave.
Suite A104
Oak Ridge, TN 37830
Tel: (865) 425-0301, Ext. 160
Fax: (865) 425-0268
http://www.roboticcontrol.com
 
We have applied ball screws in a similar application where limit switches were not used and incremental feedback was derived from the
resolver based servo motor. We managed the "powering into the end-stops" in this manner. The servo amplifier was operated in torque mode and whenever you powered up the system, the servo controller voltage output (DAC Limit) was clamped to a low value (15%) as was the JOG speed. This prevented the operator from being able to damage the unit. Also mechanical stops on the load were placed such that the ball screw nut/screw were unable to bottom out against mechanical components
in the actuator. Homing is accomplished by powering into both mechanical stops (forward first, then reverse) and recording the encoder
position at each end of travel. Then software limits inside this envelope were placed into effect, full DAC limit restored and JOG speeds
set to full speed - also at this time, we set a bit allowing for automatic mode operation. Software limits (position and decel rate) were set aggressive enough to bring the axis safely to a stop. For reference, these were 1.5" ball screws with 140msec moves of 3 inches or less and we have several dozen systems running now 7x24 year round with no difficulties.

Ken Brown
Applied Motion Systems, Inc.
http://www.kinemation.com
 
Top