Flux vector VFD versus simple VFD?

[Andrew TinsleyParticipant @andrewtinsley63637]

[Andrew TinsleyParticipant @andrewtinsley63637]

So what is the difference in operation between the two styles of VFD? I am told that Flux vector VFDs have more torque at reduced revs. How does this work in practice?

Regards,

Andrew.

[Neil WyattModerator @neilwyatt]

It uses feedback from the motor to improve the control. This allows them to use higher voltage/current at low revs to boost torque without risk of overspeeding.

Even the inexpensive IMO Cub I have uses 'simplified vector control'.

Neil

[Mike PooleParticipant @mikepoole82104]

Flux vector control needs the motor to be turning to be able to use the feedback it gleans by making some accurate measurements from the motor and some powerful computation to determine rotor position. A squirrel cage motor can deliver 100% torque at a standstill if you fit an encoder that gives the control system the rotor position. Sometimes it is called sensorless flux vector control which confirms it has no rotor position sensor. Although the flux vector drives are usually more expensive I would choose one if possible, the difference is remarkable if you have a drive that can run in either mode, the low speed performance is very good on the vector drives but it is still worth getting in the right ball park with the belts or gears if you want a very low speed.

Mike

Edited By Mike Poole on 20/02/2018 22:28:30

[John HaineParticipant @johnhaine32865]

To simplify…A typical induction motor generates max torque at around 90% of synchronous speed. So say 2700 rpm for a 2 pole motor. At that speed the rotor actually sees a rotating magnetic field at 300 rpm, or 5 Hz, the slip frequency. If you fed the stator with 5 Hz and controlled the current to generate the same field and the rotor was stationary the field would look exactly the same to the rotor and it would generate the same torque. Since the currents will be the same as when running normally but the fan is stationary a separate cooling fan may be needed.

[not done it yetParticipant @notdoneityet]

Power is calculated from torque and rpm. To avoid halving the power, by running at half speed, the torque needs to double to maintain the same output.

So, in practise, the flux vectored VFD provides more power at slower speeds than would otherwise be provided. It may not give constant power, but sure helps out.

[Anonymous]

Let's start with the basics, the torque produced by an induction motor is primarily determined by the phase current. At base speed, usually associated with a frequency of 50Hz in the UK, the motor is designed to take rated current, and hence produce rated torque, at the rated voltage.

As the frequency decreases the phase currents will rise for a fixed voltage primarily due to lower back emfs. So, to keep at the rated current the applied voltage needs to be reduced as the frequency reduces. Simple, open loop, VFDs normally use a fixed V/f curve. This works quite well, but tends to go wrong at low frequencies, and voltages, where winding resistance is of greater importance.

Vector, or field oriented, control is different. Essentially the three phase rotating vector is transformed to a quasi-stationary two phase vector via the Clarke/Park transform. The resulting current vector has two components D (direct) and Q (quadrature). The D current represents field flux linkage and the Q current represents torque. So torque, and other motor parameters can be explicitly controlled. And features like torque boost at low speed and slowing the motor under overload conditions to prevent stalling are possible.

Vector control can be open loop (no sensors) or can use current sensors. A limitation of open loop vector control is that the motor must be rotating, ie, there is a minimum speed for control. This isn't normally a problem for machine drives. Vector control with sensors can provide full torque at zero speed. This is important for electric vehicle drives, where I cut my teeth on VFD design. It's the equivalent of slipping the clutch on a hill start.

I'm mystified as to how incorrect currents/voltages could cause an induction motor to overspeed. Surely rotation speed is driven solely by the frequency, plus an allowance of a few percent for slip?

Andrew

[Neil WyattModerator @neilwyatt]

Posted by Mike Poole on 20/02/2018 22:27:14:

Flux vector control needs the motor to be turning to be able to use the feedback it gleans by making some accurate measurements from the motor and some powerful computation to determine rotor position.

Not entirely true, the fact there's no back-emf is enough to tell the VFD the rotor isn't turning. It can then apply an increasing starting current until it starts to turn.

Neil

[MuzzerParticipant @muzzer]

The better performance of the vector drives comes at a cost. For instance the Yaskawa J1000 drives are "simple" control (ie constant V/f) and the V1000 drives are their vector brothers. They look almost identical but inside you will find that the V1000 models have an extra board with an FPGA or DSP to handle the high speed, real time computation – and the software to make it happen. The cost difference is about £40 for a 1.5kW version. As mentioned above, torque is a function of the quadrature current, so they will be using the phase current measurements as the raw data along with the nameplate parameters you enter when you set the drive up initially.

Murray

[Andrew TinsleyParticipant @andrewtinsley63637]

Thanks gentlemen,

That really was a very informative series of descriptions. All is now clear. Interestingly, when I have been browsing VFDs with a view to purchase in the mid term, there appears to be hardly any mention of a VFD being a flux vector or a simple version, in the sales summary. One has to delve into the specifications to find that info. Strange, as I would have thought this would have been shouted from the house tops!

Thanks everyone,

Andrew.

[Neil WyattModerator @neilwyatt]

Posted by Muzzer on 21/02/2018 13:48:28:

The better performance of the vector drives comes at a cost. For instance the Yaskawa J1000 drives are "simple" control (ie constant V/f) and the V1000 drives are their vector brothers. They look almost identical but inside you will find that the V1000 models have an extra board with an FPGA or DSP to handle the high speed, real time computation – and the software to make it happen. The cost difference is about £40 for a 1.5kW version. As mentioned above, torque is a function of the quadrature current, so they will be using the phase current measurements as the raw data along with the nameplate parameters you enter when you set the drive up initially.

Murray

Don't confuse with 'simplified vector' which is more than 'simple V/F' but not as sophisticated as a full implementation of vector control.

Neil

Edited By Neil Wyatt on 21/02/2018 14:58:19

[Author]

Posts