AC vs DC servo motor in high acceleration application

J

Thread Starter

JonKaldi

Hi all,

I am designing a positioning system using a servo motor to move my load. The motor is to drive approximately I_load = 0.07 kg*m^2 load from a stop position, only 1 degree increment in 3.5ms.

This will be a stepping system in that sense it will step 1 degree every 3.5 ms. The motor should accelerate (linearly) the load from 0 to 11.3 rad/s in the first 1.75ms and then decelerate from this 11.3 rad/s to 0 in 1.75ms.

I have found out that I need a motor that can deliver approximately 20-25 Nm continuous torque, where the max velocity of the motor will be bit more than 100 RPM. I really only need 10 Nm motor, but I want to be on the safe side, only driving the motor at half its capabilities. This torque is calculated with gearing ratio of 1/10.

I don't want a stepper motor since it has various drawbacks such as vibration of the load while stepping (i need smooth movement), and also I want feedback possibility to control the position precisely, and not micro-stepping.

The question is, do I need DC servo motor or an AC servo motor? and why?

I have heard that AC motors have low torque at start, where as DC do not and therefor DC would suit me better, is this true?

Do you suggest any specific motors that would be suitable?
 
K

Ken Emmons Jr.

What you need is a 3 phase permanent magnet motor with encoder, and a servo drive that can control it. Depending on your requirements basic motion control can be built into the drive/amp, or you have to use a separate motion controller. If all you want to do is increment one degree I'd go with a basic drive with intelligence in it.

Keep in mind that depending on your duty cycle you can often deliver 300% of the rated torque with such motors. The only thing you have to be careful of is that since torque equals current (i.e. through the torque constant) your power dissipation goes up as a square of current (i.e. i^2R). This is where the duty cycle comes in.

KEJR
 
W

William Sturm

The terms AC and DC servo are widely misused and don't have much distinction anymore. A DC servo used to mean a brush style permananet magnet motor. It can also mean a trapezoidally commutated brushless permananet magnet motor. AC servo has been applied to sinousoidally commutated brushless permananet magnet motors and possibly also to AC induction motors with a closed loop position control.

If you are using a permananet magnet brushless motor, which most servos are these days, you probably want sinusoidal commutation for smooth low speed response. A trapezoidally commutated motor is simpler to drive, but will have a torque ripple at low speeds. A permananet magnet motor has full torque at zero speed. An AC induction motor has reduced torque control at very low speeds, they are also not commonly used as a servo motor.

I hope this clears the air a little,

Bill Sturm
 
Thank you guys so much for your reply.

As I understand both of you, I should use a (brushless) DC permanent magnet (3 poles) motor.

The duty cycle is 70%, the movement takes 3.5ms and the cycle is 5ms, so according what you, Ken Emmons Jr., is saying I could force the motor to deliver a bit more torque than its rated continuous torque since I am "only" using 70% instead of 100% duty cycle. I hope I understand correctly?

The idea is to just buy a motor with a motor controller/driver which (hopefully) would just plug into 230Vac for power, and could communicate with a PC with some kind of a bus (CAN, USB etc.) for control and encoder feedback.
Can you suggest some vendors/manufactures to me? It would be nice, if the same company would provide gearing as well (~1/10 ratio).

Also I want to hear if other gifted people would like to confirm or not confirm the statements made here about which kind of motor I should select?
 
W

William Sturm

JK> As I understand both of you, I should use a (brushless) DC permanent magnet (3 poles) motor.

Yes, most likely.

JK> The duty cycle is 70%, the movement takes 3.5ms and the cycle is 5ms, so according what you, Ken Emmons Jr., is saying I could force the motor to deliver a bit more torque than its rated continuous torque since I am "only" using 70% instead of 100% duty cycle. I hope I understand correctly?

Yes, but plan for some extra margin. Do not plan to drive the motor to 100% average power under normal long term conditions. The cont. rating of a servo motor is specified to keep the motor temperature within limits. The peak rating is for instantaneous demand.

It is not easy to recommend a servo system w/o a lot more info. Here are a few reasonably well known sources that sell servo gear motors and drives.

http://www.exlar.com/prod_SLG_oview.html
http://www.baysidemotion.com/web/bmghome.nsf/?Open

I have always liked Kollmorgen drives and motors. They sell gearboxes too, but possibly more interesting for your application is their direct drive rotary motors. They are high torque motors with low base speed (< 800 RPM), meant to eliminate the need for a gearbox. (no backlash is a great thing for precision apps)

http://www.danahermotion.com/websit...tors/brushless_direct_drive_rotary_motors.php

Bill Sturm
Abbeytronics LLC
 
I wouldn't spend any time worrying about whether it is an AC servo or a modern brushless DC servo. The difference between those two characteristics can be much less significant than other design differences that you may find between servos from different companies.

Get in contact with some reliable servo dealers who know what they are talking about, and discuss your application with them. They should be able to recommend which of their models will suit your application the best. You will get much better advice that way than you would by listening to opinions here. They should also be able to supply compatible gearboxes and couplings.
 
K

Ken Emmons Jr.

Hello,

This is correct. Although a 70% duty cycle is pretty busy, it gets you a 30% margin which is nice.

If you are a halfway decent programmer and like the power of a programming language I recommend Elmo amplifiers. They have a high level language that you can program them in, perform well, and also allow you to communicate to a host via input/output signals, RS-232, and CAN (Not sure but it is probably CanOpen protocol).

If you want absolute control and performance Delta Tau motion is your safest bet. They do have integrated drives, but I have had best luck with the standalone motion controllers hooked up in torque mode.

Unless you have a very complicated motion profile I'd look at the Elmo
amps.

~Ken
 
roshu, I appreciate the link. More information is usually a good thing. In this case, however, the data is comically bad.

AC servos are "Low power output of about 0.5 W to 100 W."? No. I don't think I've ever used an AC servo less than 1 <b>kilowatt</b>. But that's just the industry I'm in. There are AC servos in the <b>hundreds</b> of kilowatts.

Where things really break down is commutation. I'm not really complaining about the badly mangled English in general. However, "commentator" is flat out wrong. The correct word is "commutator." That's just the first of the problems. Next, we can now see that the author is speaking about <b>brushed</b> DC servos. The comparison of DC brushed vs AC servo is worth doing, but nothing makes it clear that that is what's being compared. This issue was commented on by several people earlier in this thread. Finally, AC servos also have a commutator, it's just that it's an electronic one. It's absolutely true that brushes have to be replaced and electronic ones rarely fail. But this entry in the chart is horribly misleading.

Also, the idea that AC servos have no radio frequency noise is laughable. In most applications, managing noise is critical. I can remember several applications off-hand where noise caused me some severe headaches, one of which involved a couple of weeks of man-hours to find and fix. (Long story short; we had to use a completely separate 24VDC power supply for the drives auxiliary functions, because it would inject noise onto the line.)

Again, I appreciate you digging up this link. Still, the information there is wrong. (I know, I know! Someone is wrong on the Internet, and somehow me commenting on it will improve things. Can't help myself.)

-James Ingraham
Sage Automation, Inc.
 
Top