Problem with Temposonic/Servo oscillations

lonewolf454

Member
Join Date
Oct 2009
Location
Georgia
Posts
5
I have a preexisting condition on a servovalve controlled cylinder when it is at endpoint it will oscillate. We have Temposonic (blue) probe in cylinder and 1771 QB module. I have renulled valve and also checked for loose connections. PID gains appear not to be set too high P 0.4, I 0.5, D lowered it from 0.5 to 0.1 and also added deadband of 1" - and it appears to have helped, but still when machine is first started it oscillates for first 30 minutes. I can probably change servo as we have in past and problem will go away - temporarily. Something is causing this problem but I am unsure what else I need to check. There are 3 other servovalve controlled cylinders on same system and they are ok, one is on same QB module. Any suggestions?
 
This is only a suggestion that you can give a try
You may have to tune your P I D once again. I think that currently the system is stable but oscillatory in the transient stage. The correct P I D values will depends on your system characteristics and it is not possible to tell how you should change it from this end. Try tuning it and you may not required to use the deadband
 
Have you considered that you may find you have air trapped inside the hydraulic cylinder, as the air is compressable it can cause movement when control valves operate.
 
Does the oscillation occur at both ends of travel or just one? Does the cylinder oscillate if you stop it in the middle of travel? Did you have a failure analysis performed on the servo valve you replaced last time?

Since you say the oscillation only happens for the first 30 minutes that seems to indicate that time or temperature is changing something in your system.

Keith
 
Since you say the oscillation only happens for the first 30 minutes that seems to indicate that time or temperature is changing something in your system.

Keith

when it is at endpoint it will oscillate.

lonewolf454,

This may be nothing like what is happening to you. Cylinder extended when cold would oscillate. The problem was very similar to Alaric's dilemma in this post, but his is retracted mine was extended.
http://www.plctalk.net/qanda/showthread.php?t=42424&highlight=cylinder+piston

When the cylinder was warm it expanded (grew longer) and the setpoint was not at the very endcap. Apparently the cylinder grew more than the tempo rod, when the cylinder was cold the setpoint was right at the endcap, the hydraulic fluid was restricted and the integrator would wind up. When it finally did move it would overshoot. The solution for my problem was to change the setpoint about 15 thousandths shorter.
 
I have seen it do it entire travel, somewhat. After I retuned the PID as stated and added deadband (which I have never had to use on a hyd system) it now only happens initially at startup and goes away, I am only noticing it at endpoint where cyl is extended, not retracted, not while travelling either.

Gil47> But I am not having problem with other cyl on system, if it was trapped air, seems like it would be less consistent, more erratic.

Mildrone> Sounds like a good idea, I will try when time permits.

Would seal wear cause this? (when oil warms it swells seals??)

One of our other techs said our hydraulics people told him it is sometimes cause by a bad bladder, I don't see that either, only prob with one cyl/servo.
 
Last edited:
Very little hydro experience.
Piston good, PID good, fails when cold. No bubbles.
My thought is a restriction, same word used in Vaughn's post. Bad valve, tight bend in a line causing restriction, clogged screen filter.
 
Would seal wear cause this?

Yes, also if you have a bad section of tube. The seal may be good for most of the tube, but sometimes not good enough in other high wear areas of the tube.

One of our other techs said our hydraulics people told him it is sometimes cause by a bad bladder, I don't see that either, only prob with one cyl/servo.

If you only have one accumulator then I would say yes. If you have many accumulators then a problem usually shows up during high oil flow situations.
 
Last edited:
Does the bad actuator system oscillate when the other actuators are not moving? In some systems the motion of other actuators causes pressure changes that affects the other actuators.

The viscosity of oil and bulk modulus changes as the system gets warmer. Saw mills often warm up their hydraulic systems 15 minutes before startup by cycling the actuators and warming up the oil. Warming up becomes more important as the weather gets colder.
 
Does the bad actuator system oscillate when the other actuators are not moving? In some systems the motion of other actuators causes pressure changes that affects the other actuators.

Yes it moves whenever you first start system. It is a sawmill in fact and operator doesnt like to spend much time warming system. We had this problem on same positioner in the past and have cured it by changing servo. My problem is it comes back after servo gets some wear. We change approx 6-8 a year on the 4 cyl, this seems to much to me. Our maint supv. says hydraulic analysis always comes back good. I just want to fix actual problem instead of changing servos to conceal it.
 
It is a sawmill in fact and operator doesnt like to spend much time warming system.

Tank heater? Element burned out? Ect.

I had one system without a tank heater. But I did have a PLC with a real time clock. Also this system had Do3 and a Do5 bang bang valves for speed control of a single cylinder. All I had to do was program the Do3 to extend and the Do5 to retract this left the cylinder right where it was but generated heat through the flow controls on the bang bang valves. I guess you know what to do with the time clock right?
 
Last edited:
What happens if you turn off the integrator gain when starting? Does the actuator still oscillate? How fast ( frequency ), how far ( amplitude )?

You should have a warm up program in the PLC that simply moves the actuators back and forth before production starts if you want optimum performance from your system at the beginning of production.

There length of the hydraulic lines going to the problem actuator could make a difference but if the four actuators are on the same machine the line lengths should be similar. Hopefully the valve is mounted on the cylinder.

These kinds of problems are hard to diagnose without pressure gages and being there to see the layout.

Usually these problems are caused by valves with a dead band or closed center valves but these shouldn't be used in sawmill hydraulic servo systems. You might check though.
 
Originally posted by Peter Nachtwey:

Usually these problems are caused by valves with a dead band or closed center valves ...

How would this cause a problem that disappears after some period of time? I would think the problem would stay the same if it were valve design.

Keith
 
I agree with kamages, doesn't seem like it should just be initially, it would be erratic all of the time.

We are not using tank heater yet, it is still warm here, and I doubt if the 100gal? reservoir actually "warms" up much in the 15-30 minutes it takes for prob to go away. We can "work" (or warm up) the cyl for a few minutes and it helps.

I have thought about a warm up prog to move all setworks for a period of time, we dont have that in our program. Would think it would help though.

As I stated before though, it is this one cyl giving us trouble, has been having these sort of issues since I have been here (2 yrs) and can fix usually by changing (Atchley) servo valve, but always comes back after a short while.

I did try resetting PID loop P .004 I and D 70% of P, didnt help, tried swapping QB, same prob, tried different probe to see if I was getting misinformation back from tempo, same prob. Maybe next week...

Have a great weekend.
 

Similar Topics

Hi We're having issues with an FX1N-1DA-BD Analog output module. We're using an metallic contamination sensor MCS 1000 to detect particles. This...
Replies
1
Views
60
Hi everyone, I'm trying to simulate any program in control expert and see a register in Modscan32 or any software to do that (Like ModbusPoll). I...
Replies
0
Views
62
Hi all. I´ve got a little problem with my FT View SE alarm window. On my PLC program, I have 200 SINT for alarms, in which every bit is for one...
Replies
2
Views
134
Hi all, Im stuck trying to support a client with Motorola ACE.. I was able to upload all of the other sites to STS but one. Im working remotely...
Replies
0
Views
77
Hello I am writing to request assistance with obtaining a specific GSD file for our Mitsubishi inverter type A800-E, which is connected to our...
Replies
10
Views
227
Back
Top Bottom