Developing a robotic app 7

[dpenney]

I am developing a robotic app which requires quick changes
in direction and speed.

For reliability, I want to avoid brushes and so it seems my
choice is either 3-phase induction (squirrel cage) or
brushless DC.

The app will require approximately 3/4hp & 0-2400rpm.
My power will be wall current (120V AC) and I will use a
microcontroller with appropriate inverter for either
approach. It would be great if the motor could reverse in
100-200ms. Position control does not need to be precise
so a stepper or servo (as I think is usually used for this
type of app) is not warranted.

Which motor type would be best for this? If both can
generally meet the performance requirements, does 3-phase
have a cost advantage over brushless DC?

Thanks,
dpenney

 

[cbarn24050]

HI, brushless motors are the way to go, much more expensive.

 

[dpenney]

Thanks for your reply.

Can you elaborate on why DC is better for this than 3-phase
AC (both are brushless)? Obviously, I would prefer to get
by with 3-phase AC if it were possible, for the lower cost.

 

[Steventyj]

Have you considered changing direction though some kind of a mechanical system.

 

[jbartos]

Suggestion: Please, would you describe the motor duty that the robotic application requires?
The squirrel-cage induction motors are normally not designed for frequent reversals. It would need a heavy duty service inducton motor product. It is more expensive and heavier to dissipate the heat.

 

[ElMimino]

If you are absolutely hot for reliability, you can try something called a linear induction motor. These are great for reciprocating applications, and they make them in a rotary orientation.


Rich Schmidt

 

[SERVOCAM]

The advantage your AC Induction motor will have is cost. The disadvantage is the ability to quickly accelerate or decelerate the load. You could do closed-loop vector (pretty much sevo control of AC induction motor) to over come this. The Brushless servo will give you high speeds, quick accels/decels, and better control.

Some more application detail would be nice.

ElMimino, sorry but I got a little laugh about your comment on making a rotary orientation linear induction motor. Isn't that just a regular AC Induction motor (or are you talking about a ring-track bearing and a round aluminum reaction plate?).

If you need too reverse cycles at a high frequency, then a mechanical system (as Steventyj was stating) may be the way to go - Crank or cam type sytem.

Cameron Anderson - Sales & Applications Engineer
Aerotech, Inc. -

http://www.aerotech.com

"Dedicated to the Science of Motion"

 

[dpenney]

The app is similar to an animatronic display (like the
dinasour types, but much faster moving) that is interactive.
So there may be slow movement in one direction, but then an
abrupt reverse and/or speed increase (or fast movement and
speed decrease), in response to some stimulus. Several
motors will be needed for several axis of motion and much of
the movement will be random, non-repetitive, controlled by
a PC.

So, does it still sound like I can get the performance I
need (at the lowest cost) with the 3-phase AC induction
motor (but heavy-duty type per jbartos input), using servo
control?

Thanks for all your responses.

 

[Steventyj]

Are you going to need to be able to provide holding torque or will this thing be in constant motion with a predictable rest state when it is turned off? It may be worth looking into motors with electric brakes, this may also be a safety issue - eg power outage doesn't cause the creation to hit someone on its way to its natural resting location.

 

[automatic2]

No, it is unlikely, due to low/zero holding power. Unless you use your motor to drive a hydraulic pump. The hydraulics are then controlled to drive linear actuators, which in most cases, emmulate limb motion far better than rotary electro actuators. A very interesting valving system that was produced years ago by the North American company utilized a hydraulic nozzle that gave excellent response to analog input. Far better in my opinion than current electro servo type valving. If this valving was driven by a voice coil, extreme motion control could be had with analog signals.

 

[moturcu]

The difference between brushless DC and induction motor becomes sharper when we talk about efficiency and reliability issues. Other performance requirement will be satisfied one way or other as long as they are controlled properly. For example frequent direction reversing related temperature rise can be overcome by dynamic braking or regenerative operation.

As far as the reliability concerns, induction beat brushless DC motor. On the other hand, efficiency will be higher in brushless DC than in Induction motor which is an advantage in the long run.

One more thing I need to add is the physical size comparison. Brushless DC will be smaller than induction motor for the same power level.

 

[dpenney]

Yes, it will require holding torque. However, I had hoped that the brushless motors in my app could perform similarly to a variable speed power hand drill. I understand a drill usually uses a brushed (Universal) motor and it seems to have adequate holding torque (light trigger press will hold steady against counter-torque).

Can an AC induction motor perform the same way?
Or does it require an impractical amount of current to do so?

automatic2- I considered a hydraulic system but decided to avoid it due to the high pressure required for the faster movements. Also, hydraulics are even farther out of my area of expertise!

If AC induction can't provide holding torque then I think I would sacrifice reliability and resort to using universal motors (brushed). But then I think there are additional problems from the high voltages of the quick direction changes, not to mention safety issues if it sparks.

moturcu - I would like to learn more on how to beat the heat problems by dynamic braking or regenerative operation. Can you point me to any web resources on it? Is it as simple as shunting the inductive kickback voltage into a capacitor?

I don't think the poorer efficiency and larger size of AC induction vs. BLDC is a problem for my app. Efficiency means less power required, and less heat, which is beneficial but I guess wouldn't outweigh the cost advantage. Any other efficiency implications?

Thanks,
dpenney

 

[moturcu]

If you do not care the efficiency so much, then going with the dynamic brake is much cheaper and simpler than regeneration which requires controlled AC/DC interface.
Whereas Dynamic brake approach is to connect proper resistance across DC bus during reversal.

Speed reversal takes place in two steps:decelaration and accelation to other direction. During decelaration stage, motor acts as generator, and pumps the energy to DC bus capacitor. This increases DC bus voltage. A simple controller, connects resistor during this high time of the DC bus to absorb energy and keep the voltage level constant.

Following web site may be helpful.

http://www.reliance.com/prodserv/standriv/appnotes/d7733.pdf

 

[SERVOCAM]

I'm guessing that you do not want to openly give all the details of your appliction, so I suggest taking a look at Emerson Control Techniques -

http://www.emersonct.com

They make all types of motors, Brushless Servo, AC Induction, Regen, Open-Loop Vecotor, Closed-Loop Vector. Talk to one of their Applications Engineers and they can help you determine what control system will be best for your application.

If linear actuators can help do the movements, maybe even like through a Stewart Model, take a look at Exlar,

http://www.exlar.com

for Electro-Mechanical actuators.

Good Luck!


Cameron Anderson - Sales & Applications Engineer
Aerotech, Inc. -

http://www.aerotech.com

"Dedicated to the Science of Motion"

 

[jbartos]

Suggestion: The permanent magnet motors are not exactly designed for frequent abrupt stops. It affects the permanent magnet magnetization.

 

[dpenney]

Thanks all, you've been very helpful.

 

[rlounsbu]

While you say precise position control is not needed, a stepper motor should still be an option for your quick reversal and position holding application. Better yet, for smooth motion (variable speed and torque) I would go with Brushless DC in the form of a 3-phase Permanent Magnet motor driven by a inverter using Field Oriented Contol . You get the best of all motors in one that can be used for all types of applications. There will be NO problem with de-magnetization of the rotor magnets if you use a control system that keeps the current on the q_axis and off the d_axis. Theoretically you can demagnitize, but in fact it would be pretting difficult to do even if you try. You can do fast braking and reversal by shifting the stator field 180 deg to reverse torque.

 

[jbartos]

Suggestion: Visit

http://www.hydrofloamerica.com/products/dcdrive.html

for an electronic overload protection in DC SCR Variable Speed Drive that prevents the permanent magnet motor demagnetization. If the electronic protection is somewhat off or malfunctions, the permanent magnet motor will eventually become demagnetized to a noticeable extent.

 

[moturcu]

Suggestion for previous posting:

Current control is an inherent feature in a permanent magnet DC motor based system. Controller and current regulator will always make sure current will not exceed certain prescribed level while they shape the current as sinusoidal or trapezoidal. So we should not worry about the demagnetization as long as current regulator is there.