[refuser]

I am wondering if there is anything that prevents the use of a ordinary industrial induction motor as a servo motor, provided the appropriate control?

I realize that it would not be optimal but perhaps for a hobby/diy project it could be a cheap way of getting a lot of power?

/Einar

[Al_The_Man]

There are some VFD's that can come close, and used in elevator and crane hoist control, and also spindles can position via an encoder for tool changing, but it definitely requires the feedback (encoder) option, but they cannot obtain true synchronism as a AC servo does, by the time you have gone through that, a servo with gearing is going to be the best of the two for precise PID loop control, IMO.
Al.

[refuser]

The thing is that I'm studying electrical engineering and are going to take classes like "Non-Linear Control and Servo Systems" and "Power Electronics - Devices, Converters, Control and Applications" as a part of my masters and I would like to try to use my knowledge from these classes to control an ordinary induction motor with a self made drive.

I just can't find the difference between a AC servo (induction) and an ordinary induction motor except for maybe more poles and lower inertia in a servo.

What would happen for example if I smacked a AC servo drive on an IM (with encoder) instead of using a VFD?

[Al_The_Man]

First the AC servo is not usually induction, it is P.M. Permanent Magnet field which make it possible to be synchronous with the applied frequency, an induction motor never can be.
I have not had success with PM motors on VFD's although there are such as Hitachi that claim a BLDC version, but when I inquired they said it had to be tailored to the motor in question.
Al.

[refuser]

Yet there exists induction servo motors, like the ABB HDP induction servo motors for example.

I realize that driving a PMSM with a VFD is not a good idea in general since in a IM the field is rotating much faster than the rotor as opposed to the PMSM where the rotor is rotating synchronously with the field. So naturally the drives will be different.

The question remains though. If an encoder was mounted to an ordinary IM and a suitable servodrive was used. Would it function as a servo system?

[Al_The_Man]

Using a I.M. on a AC servo drive is probabally not going to work in a servo application, the PID loop is never going to be tuned due to the non-synchronism of the motor?
A VFD is designed specifically for the characteristics of the Induction motor.
You will find alot of information on various drive requirement design on the MicroChip site in application notes.
This is based on my personal experience, there may be more information out there to contradict it, but I think IM servo would be more prevalent if it were practical?
Al.

[refuser]

But there are servo drives that are specifically designed for IMs as well. As there exists such servo systems, using induction ac servo motors.
From the ABB HDP Manual: "The majority of HDP Servomotors are equipped with rotors done by pressure die cast aluminum, of the squirrel cage type."

So there are servo systems using the induction motor principle.

My question still remains. If an ordinary IM was fitted with an encoder and driven with a servo drive specifically designed for induction motor servo systems, like the ABB HDP for example.
What would the difference be from using an ABB HDP servo motor?

I don't mean to be nagging or disrespectful, I just thing we are talking about different things =)

[Caprirs]

But there are servo drives that are specifically designed for IMs as well. As there exists such servo systems, using induction ac servo motors.
From the ABB HDP Manual: "The majority of HDP Servomotors are equipped with rotors done by pressure die cast aluminum, of the squirrel cage type."

So there are servo systems using the induction motor principle.

My question still remains. If an ordinary IM was fitted with an encoder and driven with a servo drive specifically designed for induction motor servo systems, like the ABB HDP for example.
What would the difference be from using an ABB HDP servo motor?

I don't mean to be nagging or disrespectful, I just thing we are talking about different things =)

I am not an electrical engineer, just a mere mechanical engineer so I am speculating on this.

Tuning the system is more difficult with the IM compared to the PM motor. In the motion control of mills and lathes, the servo must be tuned to work at 1000-10000 rpm for rapid positioning and also at < 1 rpm for precision positioning. Also, a mill moving the X axis would also be required to hold the Y & Z at 0 rpm, potentially for very long periods of time.

The practicality likely depends on the level of accuracy and control that is desired in the final result. A mill trying to create smooth finishes for the optical industry will have different requirements than the luggage conveyor belt at the airport.

[underthetire]

Could you get something to work, sure. Would it be good enough to run an axis on a cnc? No. A cnc requires all axis motors to interpolate with a set amount of lag or servo error to each other, in order to cut angles, circles, etc.. less poles makes it more than difficult to tune for this, and at low speed will not produce a smooth rotation, translating to flat segments in a circle. Now if this is for a different application such as pick and place machines where slow feed and circularity doesn't matter, it may work ok.

[refuser]

So what is it that makes an ac induction servo motor work that an ordinary induction motor doesn't have? Number of poles?

[Al_The_Man]

Looking at the ABB manuals they do have some impressive results, and also go to 6 poles for motors, the typical AC servo is around 6-8 pole, the higher the pole count, the better low rpm resolution, but frequency has to increase for a given rpm.
In a IM the stator winding is responsible for the rotor field as opposed to P.M. rotor in a AC/BLDC servo
I would like to see some performance figures of actual examples in a CNC application, if this has been done?
Al.

[refuser]

That would indeed be very interesting =)

I just found an PHD thesis on IM vs PMSM in motion control. It looks like its sponsored by ABB and they are using a modified rotor to make a PMSM into an IM. Maybe I'll find my answers there =)

But from what I can gather, an ordinary IM would probably overheat in any kind of servo application.

[Al_The_Man]

But from what I can gather, an ordinary IM would probably overheat in any kind of servo application.

I would think it is a given that fan cooling would be needed, this is typically standard on CNC spindles which posses V.F. drives.
Also probably goes to the fact mentioned, the stator winding provides the rotor current?
Al.

[mike_Kilroy]

So what is it that makes an ac induction servo motor work that an ordinary induction motor doesn't have? Number of poles?

no. no of poles is a non issue. time to step into this discussion with, say, 'another perspective,' to help you understand.....

first a couple facts:

1) why will a synchronous motor not run well on a v/hz vfd? because a vfd is a simple voltage output drive that varies freq along a given vhz curve: command a speed & vfd outputs corresponding voltage. simple. but recall a syncro motor has no slip so if you load it to more torque than it can put it at any given speed (freq) it simple falls out of synch and stops. no 'give.' so , no, a pm motor will not work well on most vfd drives. If hitachi or others say they can then they have upped the anti and made the algorithms account for no slip - hence the comment in previous post that 'it needs motor parameters ' or some such comment. if it is a vector drive with correct algoritms it can run a pm motor.
2) why "can't" an off the shelf unmodified IM run as well as a servo?
- will overheat at rated torque at low speed
- it has much higher inertia
- that's all. nothing else.

So looking at why it 'cannot' work, heat. an IM of course has no magnets so no permanent magnetic field so has to make one. it does it with current in the windings. current = i^2r or heat. how does the typical IM cool itself? shaft mounted fan..... what is cooling from a shaft mounted fan? goes up by the cube of the speed. so at slow speeds it does little cooling. yet the off the shelf IM was designed cheap to cool itself - at base speed, no zero speed...

can u get around this? sure, buy the next better model - that has a separately excited fan on it so it cools good always. Cooling issue solved. now that IM motor can produce full rated torque at 0 speed all day long, just like a servo motor.

next 'issue:' inertia. As you have learned in your classes so far, T=Jw/t where T is torque, J is inertia, w is speed change, t is time to do it in. So, if your IM has 10x higher J than equiv servo motor for a given torque rating (typical), then it will take 10x as long to accel (respond) or looked at another way, it will take 10x more torque to accelerate itself compared to a typical servo motor. Can it be solved sure, buy the next better model IM motor with only 2x the inertia of the equiv servo motor. Now if you buy it with 2x the rated torque, it will accel JUST as well as the servo.

In real life applications, there are 3 types of inertia loads:
1) Jl=less than motor
2) Jl=same as motor
3) Jl= more than motor.

If you look at each application, you can see how the higher J of the IM can be a big detriment or not so much of an issue out of the gate..... since Jtotal=Jm+Jl, in case 3, if Jl is 10x more than your motor Jt=10Jm+Jm=11Jm so if the motor becomes 2x higher, it doesnt much matter to the process as Jt becomes 10Jm+2Jm=12Jm - a wash. But in case 1, Jt=.1Jm+Jm=1.1Jm so doubling the motor inertia will require 2x as much torque now. So it is important to engineer the system to make best use of this fact. When doing so, it is often possible to pick gearing to get to point where 2 or 3x more inertia on motor does not hurt too much (remember gearing reduces the inertia by the SQUARE of the ratio so games are often playable.

Lastly you must have a sophisticated drive that is much more capable than the plain jane el cheapo vfds mostly discussed on this forum. ABB DOES make a couple capable models. My 20 years experience making IM act EXACTLY like PM servos is with a brand we sell called REFU (Managed Dedicated Server Hosting & Managed VPS Hosting by LiquidWeb.
.refu-elektronik.com/en/products/converters/inverter/)

This drive must be able to produce full peak current into the IM motor windings in less than 10msec to be able to respond like a pm motor. It must have the magnetic field (Isd current) flowing ALWAYS when enabled so it can produce instant torque and not have to wait for the field to build from 0. It must have update times less than 100usec to close current, velocity, and position loops just like a servo drive.

Where you will pay $ 500 for the low cost vfd to run a 2 hp IM motor, you will pay $ 3000 for this drive.

An additional note, a lot of our Kollmorgen servo drives also run IM motors in servo equivalent performance. We have applied better design plain jane IM motors the same as servos on machines. As an example, look at black/blueMax Marathan motors - lower J, available with built in encoder, separately fan cooled.

The reason slip or non synchronousism is NO issue is the sophisticated drive takes it into account. Positioning accuracy is SAME as a PM motor. Slow speed smoothness is likewise IDENTICAL to the PM motor. REsponce is IDENTICAL to a PM motor ** of same inertia and speed/torque ratings. That is key - same as identical inertia motor. See above. Since a PM motor can almost always have significantly lower inertia than equiv speed/torque IM motor, this is why the comments you get that an IM cannot perform as well as a servo. No other reason. are there applications that IM motors will never be able to do? absolutely! there will always be cases where the higher inertia of the IM will not allow it to do the job.

Now you are armed with the data you need to understand this subject. There are 72,000 deeper facts to explain all this morein depth that you can continue to study. For instance, the why's of why IM motor will always have more inertia than its equiv PM motor, etc, etc, etc.

[refuser]

Amazing answer. Just what I needed.

I do realize that it is not an easy feat to get something good enough to compete with commercial IM servo systems. But that not the point with DIY as far as I see it, it's getting sufficiently close for substantially less.

Thanks for enlightening me,

Einar