Why are my motors so big?

[mikito]

Are you ready for a laugh?

For the past few motion control applications I've designed, a strange thing has happened.

I do a thorough characterization of my mechanical system, finding all the mass moments of inertia of components hooked up to the motor, etc. Even stiffnesses of components are sometimes considered.

I then try to size the motor to get the correct load-to-motor inertia match. I am told this should be from 1 to 10. So I size the motor.

The motor is always HUGE! I always have - like - 100X more torque than I need, though the motor and load are inertially well-matched. The shaft on these motors is bigger than 1/2 inch, but the thing it's spinning is always some small disk or whatever. I'm told by my distributor that I should always match the inertias. I even see this on my cool "Smart Motion Cheat Sheet". He says that if I don't match the inertias, I won't be able to tune the motor.

What's up? [sig]<p>Mike Pinch<br>[/sig]

 

[jbartos]

Look at the professionally matched motor and load, i.e. sets (compressor, pump, grinder, fan (even one for US$0.99)), offered by a private sector business facing tough competition. Those sets seem to be well designed, matched, efficient, light, proportionately small, etc.; especially for airplanes and ships. There are many businesses data sheets available on

http://www.thomasregister.com

to evidence things. [sig][/sig]

 

[mikito]

Thanks jbartos. I will keep an eye open.

I assume from your message that I'm not correctly sizing my motors. Any others out there please feel free to advise me on the correct way to do this. My background is mechanical engineering. [sig]<p>Mike Pinch<br>[/sig]

 

[jbartos]

You are welcome. Apparently, the manufacturers of motor-load sets are not told, they just design and test things, guarantee things (real smart considering those constants 1 to 10), and sell things. [sig][/sig]

 

[ChristianSterner]

Have you considered the effect of putting a reducer between your motor and your load? We typically couple the motor to the load via a gearbox or a set of timing pulleys. In either of these arrangements, the effective torque supplied to the load is increased by the gear ratio. More importantly, the effective inertia of the load at the motor shaft is decreased by the square of the gear ratio. For example, a 10:1 reduction will make the inertia of the load seem 100-times smaller to the motor.

Unfortunately with the addition of a reducer comes a decrease in the effective speed (by definition.) However most servomotors reach speeds (at rated torques) that allow for considerable reduction. [sig][/sig]

 

[mikito]

Christian,

Excellent point! I did put reducers in the design (planetary/epicyclic) with reductions of 3:1 in one case and 20:1 in the other. I must admit, the motor would've been Really HUGE if I hadn't. As it turned out the motor was just Huge.

I can e-mail a Kollmorgen Motioneering model of an example system to whomever wants it, to check it out.

Is it possible that maybe I'm just thinking these motors are big compared to the load/application because I'm used to seeing consumer products which are typically open-loop which don't need load-motor inertia matching? These applications might typically be sized based on torque and thermal/overheating requirements right?

Is it possible that when &quot;auto-tuning&quot; using various controllers' software packages (eg., Galil's SDK or Kollmorgen's tuning program), that the motor needs to be inertially matched, but if you're calculating PID gains using pole placement techniques then you can stick a smaller motor in the system and it'll work.

I read through Jacob Tal's (Galil Motion Control) &quot;Step by Step Design of Motion Control Systems&quot; and there's a chapter (7) which discussed motor selection/optimization based on various targets (optimum velocity, gearing, temperature, dissipation, etc.) This left me with the feeling of knowing more but also realizing that I hadn't put much of it into practice, since you'd practically have to go out and buy motors and set up a design of experiments to confirm it all.

At this point we can continue the thread, or call it quits. I suppose what I need to do is experiment more with motors. If you all want to continue this dialog great! [sig]<p>Mike Pinch<br>[/sig]

 

[jbartos]

Suggestion:
Check &quot;overview document&quot; covering Chorus Motors by Borealis at

http://www.chorusmotors.com/index.htm

Chorus will reduce the size depending on the application by >50%
by increasing the number of phases [sig][/sig]