Running Synchronous AC brushless servo (Mavilor BLS 112A & Mavilor 113A) on VFD

[karkas44]

Hello everyone,

I am new to servomotors, and there's a client I have who wants to see if it's possible to run these motors mentioned (Mavilor BLS112A & 113A) which are brushless, synchronous AC motors on a VFD. They are having problems with their maintenance staff to work with the servodrives. I am an electronic engineer but I have only worked with induction motors in my experience, I can't find if these motors are permanent magnet ones, I am not sure if I should just assume it.

These motors are supplied with sinusoidal waves. Has anybody done this? Information about the torque speed curves for the load are not available, just the ones for the motors.

http://www.mavilor.es/pdf_products/bl100_series_sc.pdf

Thanks in advance.

Regards.

 

[cswilson]

These are definitely permanent magnet brushless servo motors. If your VFD is intended for open-loop induction motor use, it is very unlikely you will be successful running these servo motors with it.

Induction motors can still generate torque when there is a large difference between the electrical frequency and the mechanical frequency (rotor speed). PM servo motors cannot. To start the motor, your VFD would need to be capable of gradually ramping up the frequency and voltage at a rate that the motor can always follow, even with its load.

Even once up to the desired frequency/speed, if you ever exceed the peak torque capability of the motor for that input (when the motor lags the input by more than 90 degrees of the cycle), you will have a complete failure of the process, and probably a violent one, as the motor will completely lose synchronicity with the input signal with no way of regaining it.

Curt Wilson
Omron Delta Tau

 

[karkas44]

Hello Curt,

Thank you very much for your response. I have found a Drive with Permanent Magnet Brushless motor capabilities, this is the drive

http://www.ab.com/en/epub/catalogs/36265/1323285/9544342/Introduction.html

As you can see it can control this kind of motors and in supports inputs for resolvers, I am still studying the manual to make sure this drive will run these specific motors effectivesly. But seeing this info about this drive, I would like to ask you, what do you think? Is there any other consideration to be taken into account with this drive? I mean besides finding the correct frame specifications an such things in the manual.

Thanks in advance.

 

[cswilson]

Yes, that drive has the capabilities to do it -- you have to make sure you have the proper voltage and current ratings, and the interface to the position feedback device you use.

Curt Wilson
Omron Delta Tau

 

[karkas44]

Hello Curt,

In the slection guide for these drives, I found a resolver-drive compatibility table that shows most of all Tamagawa resolvers, actually the resolver specs for these Mavilor motors has been very hard to find, in the specs of the motors they only give excitation voltage and frequency range; I am not sure I can just choose a drive for PM motors with resolver inputs assuming it's going to work.

Some drives I have found have a capability of controlling PM motors in open loop operation. I am trying to find resolver compaitility for them, but what would be the implications of controlling these motors in open loop configuartion? The types of control used by them (Yaskawa A1000) are voltage control method and filed-oriented current control method. Are these methods too ineffective? Is it possible to successfuly give up the use of the resolver with them?

Thanks.

 

[cswilson]

In principle, the field-oriented current control method (with no shaft sensor) for a PM brushless servo motor should be able to do what an open-loop VFD could do for an induction motor. However, the drive does need a reasonably good model of the motor to do this, so you will have to be able to enter and store key motor parameters, especially since the drive and motor would come from different vendors.

 

[mikekilroy]

I am new to servomotors, and there's a client I have who wants to see if it's possible to run these motors mentioned (Mavilor BLS112A & 113A) which are brushless, synchronous AC motors on a VFD. They are having problems with their maintenance staff to work with the servodrives

Pray tell what do gain changing from a dedicated working servo drive to another brand servo drive (yes - as soon as you add the resolver to potentially less capable AB drive you made it also a servo drive) that is not specifically set up for that brand motor????

Now if you changed servo drive to a MORE capable one, then it may make sense. But that is not the reason you gave to change. I do not think you are helping your customer doing this. I would say instead, teach them about servos so they are more comfortable with them.

http://www.KilroyWasHere<dot>com

 

[jraef]

Curtis right, that motor model is the challenge. The AB drive you linked to is capable of running servo motors, but the only models it comes with are the AB servo motor models. You can create your own, but you need a lot of motor data. If you cannot even find data on the resolver, it sounds as though other details might be difficult to get as well. The drive can auto-tune itself to the motor, and depending on the level of accuracy you are after, that may be fine. But it is a slight compromise compared to a true servo drive/motor combo.

What you could do is contact Rockwell tech support and ask if they have records of anyone else who has used that drive to run a that motor. You never know. I just ran into that situation with some old Kolmorgen motors on a specialized machine. We couldn't get any data on them because they were proprietary OEM motors, but it turned out someone else had already done it before and Rockwell had the motor model info. Worked great.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[mikekilroy]

Jraef, an aside; ask me and i can prob tell u full details on any of that brand motor.

http://www.KilroyWasHere<dot>com

 

[karkas44]

Hello everyone,

mikekilroy: You're absolutely right, actually I told them we could install the servo modules in the AB chassis and get them to work with the drives as in the other machines, but they in fact told us they didn't even want to buy the programming cable for the servodrives to work with them. In top of that, a couple suppliers already sent them quotations for the project with VFDs, I think that either they are completely sure those VFDs will run the motors from experience or they don't know what they're doing, as I said the customer doesn't have much info on the resolvers and there's almost nothing to find online, so unless the drives specifically say they can work with those resolvers it would be a gamble.

What I intend to do is find a VFD--like this Yaskawa one-- that can run those motors in open loop config and then try with the resolver option kit, if it doesn't work at least it'll do in open loop with the proper configuration. And of course, point this out with the quotation along with a full explanation on why it would be better to stay with the original servodrives.

jraef: That's great advice, I'll try and do that with the AB drives and the Yaskawa one.

Thanks a lot for your replies.

 

[mikekilroy]

Karkas, please do yourself a big favor and write your open loop VFD solution/offer with STRONG SUGGESTION it will not work but you are willing to show them...

Open loop, as Jraef told you, the vfd freq and motor freq MUST be EXACTLY the same. get out of sync ANY and the motor will simply stop. Perhaps you should try it at home before committing to your customer. I am confident you, like me and many others, will find it will not work open loop if there is ANY load on the motor.

http://www.KilroyWasHere<dot>com

 

[karkas44]

Okay okay, hold on a bit. I am starting to study this type of control for these VFDs, but why isn't it going to work in open loop if there's any load attached to the motor?

I talked to the customer and asked to check again if there's a manual or something to give me the most info they can on the motor and the resolver. Again, if I have all the necessary information on them to program the VFD why wouldn't it run the motor in open loop? Can you explain?

 

[cswilson]

If you use an open-loop VFD of the type commonly used for approximate speed control of induction motors to drive a synchronous PM servo motor, if you get even a momentary increase in load torque that takes the motor past its peak torque capability for that state for even an instant, you have completely lost control, and you cannot recover (without totally resetting the system).

A synchronous motor creates a torque per unit current proportional to the sine of the angle between the rotor magnetic field and the applied stator magnetic field. For small angles, this produces a nice stabilizing negative feedback against load changes. If the load increases, the rotor decelerates, increasing the angle difference, increasing the generating torque to compensate. If you stay within +/-30 degrees, it's even pretty linear.

But as soon as the load torque pushes the angle past 90 degrees, your nice stabilizing negative feedback turns into destabilizing positive feedback. An increase in load torque leads to a decrease in generated torque, meaning total loss of control. And even if the load torque quickly decreases again, it is extremely unlikely that the motor can accelerate again to re-establish velocity lock.

Now, the AB drive you found is not open loop in field-oriented mode, even without a shaft sensor. It uses phase voltage and current measurements as inputs to a mathematical model of the motor to calculate position and velocity. If the model is good, it can detect a slowdown that would result from increased load torque, and reduce its output frequency to compensate. But the key here is that the drive must have a good motor model to do this well, which requires entry of accurate values for multiple motor parameters such as inductances, resistances, and back EMF constants.

Curt Wilson
Omron Delta Tau

 

[itsmoked]

karkas44; This tread pertains to FOC (Field Oriented Control). I listed the parameters that are typically needed to get the model right.

thread237-390611

Keith Cress
kcress -

http://www.flaminsystems.com

 

[karkas44]

Okay, now I'm kind of disheartened. Then what's the use of the open-loop control for synchronous PM motors? It will only work with perfectly and eternally constant loads? Even with all the necessary parameters, the system will fail with a slight momentary increase in load torque?

 

[cswilson]

karkas:

There are millions of PM synchronous motors running completely open-loop in use today. They're called stepper motors. The key to their use is careful characterization of possible load torques and an adequate safety margin in provided torque. For this reason they are most commonly used in high-volume applications, where the cost of this characterization effort can be offset by the per-system savings over a large number of systems.

And to re-iterate an earlier point, "field-oriented" control methods are not open loop, even without an explicit shaft sensor.

itsmoked:

The parameter knowledge required for field-oriented control without a shaft sensor is far more involved for induction motors (your example) than it is for PM synchronous motors.

Curt Wilson
Omron Delta Tau