Servo Information Needed...


Thread Starter

Kelly, Rick


Over the last 5 or 6 years we have slowly increased the number of servo based positioning systems in this facility from none to 9 systems that include about 25 axis.

We have a well trained staff of mechanical trades people who are quite capable of trouble shooting classical mechanical systems but who have zero experience with the mechanical principals inherent
in servo based systems.

My problem is that when a servo system shuts down under a position error or torque limit our mechanical people look at us as if we come from Mars when we tell them it is a mechanical issue.
Everything they know about classical systems tell them the mechanics of the system in question are OK. They have no understanding of the mechanical issues that can effect the operation of a servo positioning system.

What I am looking for is a detailed breakdown of the skillset that a mechanic needs to have in order to trouble shoot the mechanics of a servo based positioning system. This information will allow us to build a training program to bring our current trade group up to speed. I am sure others have faced this issue.

As always... I thank everyone who replies for their input.

Best Regards... Rick Kelly

Chief Technician
Natural Cuts
Cheese Operations
Kraft Canada
(613) 537-8069 V
(613) 537-8057 F
[email protected]

Jerry Bumgarner

Have you tried contacting the servo vendor directly. They often offer training courses that are aimed directly at the technicians that have to support the eqiupment instead of the engineers that program it. I have done several like this for our customers

Steve Bailey

Jerry's suggestion is probably your best bet. Or else a course in basic physics or mechanics. Assuming that the system has been operating correctly, the only mechanical changes that affect performance are friction, inertia, and backlash. Of course, people tend to blame all problems on that component of the system that they understand the least, so the mechanical techs point their fingers at the servo drive and the E&I techs blame the drivetrain. If all parties are trained on the impact of each component on the system performance, then you stand a better chance of getting your problems diagnosed correctly and fixed sooner.

I'm sure there are a lot of good powerpoint presentations or even videos on the workings of a servo system. At PMC we have a set we use to train our sales reps with, many who only have
mechanical experience, and I'd be happy to send you some but you probably would be better off with a set from the controller house you are using.

The problems usually come from the lack of understanding of how the loop works, especially the P, I & D part. I think your people
probably fully understand inertia, overcurrent conditions, speed/torque curves and such, but the servo system looks at all these things in small increments of time that they aren't used to.
And as you probably know most of the errors are detected within those short increments and therefore they can be effected by noise or transients in the feedback loops which an all mechanical system might just steamroll over.

Once they have a feel for what might cause these problems (everything from friction to poor grounding; from emitted to conducted EMI) and how they might cause a fault, perhaps their
troubleshooting and preventive maintenance will be greatly enhanced.

My suggestion would be to contact your servo-controller company and see what they have in the way of tutorial... Maybe they'll send in an Apps Eng and give your group a mini seminar or perhaps you can just download their presentation.

Bob Close
Exec. VP
Precision MicroControl Corp

Johan Bengtsson

My guess is that the most important point is to understand the speed the servo amplifier is working at, that is understanding "high" frequency issues in the mechanical links and the fact that the servo is positioning even when it appears to be not moving.

Every axis can be thought of as a spring, ie it doesn't have to turn at both ends at the same time, the axis and everything else also means inertia, not just only the actual load.

Understanding the fact that the servo amplifier actually positions the encoder, nothing else. The encoder is normally quite stiffly coupled to the motor, that effectively means position the motor. The load is then normally coupled to the motor by a less stiff connection and resonance can occur.

And of course that applied torque directly result in acceleration not speed, and definitely not position, and what a moving or rotating mass means, but that is probably already understood by someone with good mechanical experience.

There are of course more issues but I would start with this first for mechanical staff. The control loops might be interesting too but perhaps not to begin with.

/Johan Bengtsson

P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]

After starting up one of the first big servo apps I did I realized that "Fatal Following Error" meant nothing to the machine operators / maintenance staff. I ended up writing a few routines in the controller that monitored commanded torque, warning following error, fatal following error, encoder quad loss bit, encoder modulo checks, etc. so that I could interpret the data and put up alarms that said:

"Motor Jammed" which in the controller meant - fatal following error + encoder healthy + It^2 = low "Load too high, check drivetrain lube" = fatal following error + encoder healthy + it^2 = high "Encoder failing" = It^2 = high + encoder fault, etc.

We also put in logic to watch prox. timing relative to process sequence to tell operators when a prox sequence was missed often
pinpointing problem areas for fast troubleshooting. Latching peak torques in a register that they can see on the MMI (0 - 150%) and reset at will also allows them to monitor trends in how much load it takes to run certain axes. After awhile, they get a feel for "normal"
ranges, and know when to start looking for a problem before it actually shuts down the machine.

In short, getting more "useful" information out where they can see it is a huge step forward and often clears up maintenance issues better than any amount of training.

Hope this helps . . .

Ken Brown
Applied Motion Systems, Inc.

Gerard Leemkuil

I think make a model or your control system and let people work on it. It will help to understanding the complete system.

Johan Bengtsson

I agree that this is good to do, but I don't agree it can replace training.

/Johan Bengtsson

P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]

Patrick Walsh

Most of the problems that come up after a servo machine has been in service relate to cabling and connections. If you teach them how to trouble shoot these problems, you are pretty much done. And the best method of trouble shooting them is to replace suspect cables and walk away.