Powering sync motor with VFD 3

[bklauba]

We have been working with a new PM sync motor, which has very good speed control (can tune the speed in increments of 2 RPM up to 8 KRPM) when run open loop. ( ! ! ! ) For previous apps of ours, we have used BLDC with a resolver, but have wanted to get away from using resolvers. Our app involves vibration, and when the vibration becomes severe, the data from the resolver suffers phase shifting that result in phase commutation errors, that manifest themselves as increased power to go thru the high vibration speed range. Not so with the sync motors, which can go thru the same speed range and load conditions, without increase power . . . . .

Except, we have an overall increased power and current going thru the sync motor, at ALL speeds. Using a 240 V, 5 HP A-B PowerFlex 40 drive, we see power figures of 250 - 1200 watts / 4 - 6 amps, over the speed range of 1000 - 7000 RPM. When using a BLDC motor, we will see power draw being only 50 - 800 watts, amps 1 - 4 amps, same load and speed range. (Different drive, one that is appropriate for BLDC.)

Using a conventional async motor, similar load, we do NOT see this increased amp draw from the A-B drive.

We also see the sync motor run hotter, 10 - 15 degrees C.

Are there settings on the VFD (this one or others) that can be adjusted to remove the excess amperage / power and get this to run cooler / and come closer to the efficiency we see with the BLDC ? ?

 

[Skogsgurra]

I have had two issues with PM synchronous motors and VFDs.

The first one was easily solved. It was about running the motor with cos(phi) as close to one as possible. To do that, adjust the V/Hz ratio to get as low current as possible.

The second one was more problematic, the synchronous motor has very low damping and the rotor swings back and forth with respect to the rotating magnetic field from the VFD and stator winding. An amortisseur winding (as used in larger motors) could help, but that is not something you can retrofit. If you look at the rotor with a stroboscope, you will probably be able to see the swinging (if it exists). Not easy to handle without an encoder. I have used a lead/rubber damper on some small motors with some success. It absorbs the swinging energy so swinging can't build up. But that is messy and rather difficult to apply.

So, let's hope that it is the V/Hz ratio that helps.

Good luck!

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[cswilson]

Your PM synchronous motor is fundamentally the same as the "brushless DC" motor when operating in synchronous mode. If you are truly running it open loop, you will always have significantly higher current than running a comparable motor closed-loop with feedback.

With position feedback, the controller can always orient the stator current instantaneously for maximum torque per unit current, no matter what the load.

Without feedback, the rotor will always needs to be ahead of this ideal angle. At no load, it is 90 degrees ahead, and no torque is produced. The more load torque, the more the rotor lags, and the more torque is produced. It is this physical feedback mechanism that permits the rotor to track the rotating magnetic field, but it has the dynamics of a spring with almost no damping, as Skogs pointed out.

The torque produced per unit current is proportional to the sine of the angle behind the no-load angle. In theory, you could operate at the peak of the sine curve and match the efficiency of the motor with position feedback, but in an actual application, you cannot get very close to this point, because a momentary overload would cause a complete loss of synchronization and a quite violent deceleration.

So in reality, without position feedback, you must operate well below this maximum efficiency point to maintain your margin of safety.

Skogs' point is well taken that you must optimize the V/Hz ratio of the drive to match the back EMF (generator) constant of the motor, because any torque-producing current requires voltage from the drive above the back EMF of the motor at any speed.

Curt Wilson
Omron Delta Tau

 

[Skogsgurra]

I forgot about the large Synchronous motor (6 kV 10 MW for an Asplund Refiner) that we needed to start with a "Pony VFD".

There was no encoder and the machine built up oscillations that eventually tripped the VFD thermally before the cables melted. It was a provisory set-up and the cables were, admittedly, on the thin side.

It was only when we introduced an asymmetric torque limit (-5% and +100%) that we could get the machine started at all. The asymmetry took the energy out of the backswing and that did the trick.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[MatthewDB]

The first one was easily solved. It was about running the motor with cos(phi) as close to one as possible. To do that, adjust the V/Hz ratio to get as low current as possible.

Just a dumb question here: If the V/Hz ratio is too low, does the drive attempt to hold the motor terminal voltage down, resulting in lots of VARs being sourced by the motor and drawn by the drive? You'd think the firmware programmers would be able to program that behavior out of the drive because it would never be a desirable trait.

 

[jraef]

To run a PM synchronous motor with a VFD, the drive must be capable of the complex algorithm and number crunching it takes to adapt the motor model. Even though the PowerFlex 40 is a sensorless vector drive, it's a low cost one. The PF40 drive does not have that kind of internal processing power. The PowerFlex 700 and 750 series do have that capability and can work with any IPM or SPM motor. The PF525 will shortly (it's been qualified on specific IPM and SPM motors but not all yet).


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