Flux Vector Drive Max Torque Limiting Factors + torque producing current vs Mag. Curr

[W@ttson]

Hello all,

Two quick questions about flux vector drives:

1. Suppose you have a flux vector drive that is incredibly oversized for the motor lets say you have a motor with an FLA of 100A and the drive is rated for 400A (Heavy Duty) 200% for 3 seconds. So in essence the max torque output is not drive limited.

At 0 RPM what is the motor's max torque output limit in the following two situations:
a. NEMA B motor, and b. NEMA D motor. I know that in each case the drive can produce FLT at 0 RPM, however, if you want to produce lets say 400% FLT (assume that this is more than each motors BDT), the breakdown torques naturally happen at different speeds for the two motor designs, does that matter or will the drive max out torque to the motors BDT value what ever that is?

2. The beauty of the flux vector drive is that it can make AC motors operate like DC motors. Benefit of DC motors is that the field current can be kept constant physically by applying a different source to the field than the armature. The armature current can then be manipulated so that you can have FLT at 0 RPM.

How does the flux vector drive physically keep the magnetizing current in the motor constant while only manipulating the torque producing current? Looking at the simplified circuit diagram of an induction machine, it seems that it can only raise the voltage to the stator of the machine. Is it implicitly forcing this because its maintaining the constant V/Hz ratio? Are the two parameters that it can only really change to the motor are Voltage and Hz?

For reference attached is the simplified induction machine circuit showing the components that the drive sees.

 

[Ingenieur]

The phase domain arm i contains info on
arm i, mag/ang
field i, mag/ang
speed, mech and elec
from this torque and flux variables can be extracted using a dq (Parks) transformation
they can be controlled/manipulated independently as dc quantities
then converted back to a phase/ang vector to control the pwm voltage
which in turn is related ro the arm i

the control algorithm is complicated
the key is the isolation provided by the dq transform

 

[W@ttson]

Sure I get the decoupling the two currents, however, the drive only outputs voltage at a certain freq to the stator of the motor. Its not like a DC motor where the field of the DC motor can be physically kept constant while the armature circuit has its voltage/current raised lowered.

I get how the drive can sense the current in the motor but when outputting to control the motor how does it tell the motor, only use this much for your magnetizing branch and use the rest for your Torque branch if it is only outputting voltage at a certain Hz??

The phase domain arm i contains info on
arm i, mag/ang
field i, mag/ang
speed, mech and elec
from this torque and flux variables can be extracted using a dq (Parks) transformation
they can be controlled/manipulated independently
then converted back to a phase/ang vector to control the pwm voltage
which in turn is related ro the arm i

the control algorithm is complicated
the key is the isolation provided by the dq transform

 

[Ingenieur]

Sure I get the decoupling the two currents, however, the drive only outputs voltage at a certain freq to the stator of the motor. Its not like a DC motor where the field of the DC motor can be physically kept constant while the armature circuit has its voltage/current raised lowered.

I get how the drive can sense the current in the motor but when outputting to control the motor how does it tell the motor, only use this much for your magnetizing branch and use the rest for your Torque branch if it is only outputting voltage at a certain Hz??

The phase signal has a mag and ang
this is determed by the controlled dq outputs when converted back
d ~ flux
q ~ torque
the relationship determines the field flux/arm i relationship (thru the inverters applied voltage)
and dq relationship is determind by load
just like a dc motor

 

[W@ttson]

Ok the below in red is clearing things up. That is the connection between all of the control loops within the drive and the connection to the motor.

The phase signal has a mag and ang
this is determed by the controlled dq outputs when converted back
d ~ flux
q ~ torque
the relationship determines the field flux/arm i relationship (thru the inverters applied voltage)
and dq relationship is determind by load
just like a dc motor

 

[W@ttson]

As for the NEMA B vs NEMA D design motor question, is the answer that the motor will output its respective breakdown torque no matter if in the normal NEMA torque speed curve one BDT occurs at 0 RPM and the other occurs at roughly 80% of full speed?

 

[Jraef]

As for the NEMA B vs NEMA D design motor question, is the answer that the motor will output its respective breakdown torque no matter if in the normal NEMA torque speed curve one BDT occurs at 0 RPM and the other occurs at roughly 80% of full speed?

That got a little fuzzy, but I think you are asking if the motor BDT is the upper limit if what a motor can deliver. If so, then yes. You cannot make a motor produce more torque than it is capable of producing by its inherent design limits. So can a Flux Vector drive make a motor deliver 400% of its rated torque? No, there are no standard motor designs that can deliver that much. A FVC drive can however make a motor deliver its rated BDT at any moment in time, albeit only for the amount of time inherent in the motor' thermal damage curve, usually just a few seconds at a time with adequate cooling time between instances. But this is why performance is improved; during a step change in load, the FVC can take full advantage of utilizing BDT to reaccelerate without concern for over fluxing the motor, it is in control of both aspects independently.

There is usually not much point in using a Design D motor with a VFD by the way, because the main difference between that and a Design B is that in Design D, LRT = BDT, so you have peak torque immediately at startup when used Across-The-Line. You can duplicate that with a Design B motor and a FVC drive. The only caveat is that the Design D BDT is 275% as opposed to 200-220% for a Design B, but then the slip is higher at full load, so efficiency is lower and your long term operating costs are higher. Still, if that few extra ft-lbs makes a difference, it might have a purpose. And of course if you already own the Design D, then you might as well use it, but keep in mind Design D motors are generally NOT made as "inverter duty".

 

[W@ttson]

Awesome answers everyone! Greatly Appreciated. Have a happy holiday weekend if you are in the states.

 

[mike_kilroy]

A tad more info. Real vector drives have TWO current loops: 1) for torque producing control of in phase current called Isq, & 2) for magnetizing current (90 degrees out of phase) called Isd.

Torque = 3/2 *Zp * L^2m/Lr * Yrd * Isq
where:
Zp=Pole pairs
Lm=coupling inductance
Lr=rotor inductance
Yrd=rotor flux
Isq=torque producing current

Magnetizing current is often initially set to FLC to get it built quickly. Then 10-20msec later the Isq controller brings it down to proper value, previously determined by motor parameters or direct entry in the VFD. It is held as constant as possible - at any speed and any load - from 0rpm to base speed of the motor. If running above base speed is required, then this Isq is controlled to 1/rpm - ie., linearly reduced depending on speed above base speed. So by 2x base speed, Isq is down to 1/2 normal value, 4x base speed it is down to 1/4rth normal. Run at 2x base speed, varying load (torque, aka Isq), and this magnetizing current is still held to 1/2 normal.

The end result is the total current, and its phase angle, into the motor, changes as a result of these two control loops.