Rotor vector control

[sensij1]

So it has been a few years since I've had to specify a VFD, and have been doing some homework to make sure I am up to date with the current technology. It seems that there is a new (to me) class of vector algorithm that has achieved better performance than what I think of as open loop vector control.

Emerson (Control Techniques) calls this as "Rotor Flux Control" (white paper dated 10/29/2007)
AB calls it "Force Technology" or sensorless field oriented control (white paper dated 06/2000)

Are there other AC drive manufacturers who are advertising this capability? It seems hard to tell from what they call it, but the 2nd control loop so that it uses both voltage and current feedback seems like a nice improvement.

 

[Skogsgurra]

The generic name is "Sensorless Vector Control". There's also ABBs "Direct Torque Control" or DTC - which works a little differently. But still more or less a classical vector control. Then, there's the Natural Field Orientation (NFO) that works with air gap EMF and does away with problems near zero Hz. Invertec has that.

I do not know of any drive manufacturer that doesn't have vector control nowadays. I would be surprised if there are any that don't.

Gunnar Englund

http://www.gke.org

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

 

[sensij1]

Thanks for the quick response!

OK, I see the Invertek drive and agree it fits this description. Their lower E2 line has "Voltage Vector" control, which is clearly not the same thing as their "Sensorless Vector", carried in the following lines:

Invertek (EU) - Optidrive P2
Bardac (US) - Optidrive Plus 3GV (obsolete per Invertek)


Within the sensorless vector universe, it appears that the drive can either be almost purely open loop, relying on the vector *model* of the motor to inform the control scheme, or it can be closed loop, using actual readings of the back EMF (or whatever) along with the model to make a speed feedback estimate. I'm having a very hard time figuring out which is which without looking very closely at the literature. If anyone has any other pointers towards the encoderless closed loop drive models it would be appreciated.

And yes, I understand the ABB's direct torque control is a different sort of approach altogether. Is there anyone else that has developed commercial drives based on that approach?

 

[sensij1]

Mitsubishi can be added to the list of vendors that distinguish between different types of sensorless vector control. Their E500 series has "Advanced Magnetic Flux Vector Control" which is also referred to open loop flux vector in one of their product guides. The A700 series has "Real Sensorless Vector", which significantly improved response time (300 rad/s) and 200:1 speed control.


Weg also offers two levels of control. The CFW-08 is advertised as "Sensorless Vector," "Sensorless Voltage Vector." The manual makes clear that it operates in "Voltage Vector" mode. Their higher end CFW-11 shows that there is a difference between voltage vector and "Vector Sensorless", with substantial performance improvement.

For what it is worth, I'm trying to specify a drive for a 2 HP motor running in speed control. The load is usually constant, but ideally the drive would also quickly handle the occasional impulse loads it might encounter.

After my first look, I thought a compact sensorless vector drive under $500 might get it done, but now I'm not so sure.

 

[ozmosis]

Danfoss has VVC+

http://www.danfoss.com/NR/rdonlyres...8-D57D0605E55D/0/VoltageVectorControlPlus.pdf

but as Gunnar says, most manufacturers have some high performance algorithm. They need it if they wish to compete in the VFD market where running an asynchronous motor is now the norm and PM motors, switched reluctance motors are becoming more and more common.

 

[sensij1]

Thanks for the additional information on the Danfoss drive.

So yeah, "sensorless vector" can mean way to many things to really be a differentiator on its own. At this point, the Emerson (CT) Commander SK and the Yaskawa's V1000 stand out in the <$500 range, and stepping up to ABB's ACS850 with DTC may be worth the extra cost. It looks like any of the other high performance options are >$1000, and I'm not sure enough about the application requirements to want to go there.