VFD easy question 3

[lukin1977]

Hi,

Can anyone give me an easy explanation of each of this types of controls: V/Hz, Sensorless Vector and Flux Vector

thanks,

lukin1977

 

[DickDV]

Without going into much detail, V/Hz, Sensorless Vector, and Flux Vector are increasing levels of speed control.

V/Hz simply sends a frequency to the motor to tell it to run at a specific speed and does nothing to verify that it actually does it. Typical speed error from no load to full load on the motor is one-half the motor slip speed.

Sensorless Vector controls motor speed thru the use of a calculated shaft speed feedback signal. This signal is generated thru the use of a motor model built into the drive memory at the time of commissioning. The accuracy of the motor model and the frequency of updating the calculation determine the speed error but, generally, the error will be in the range of 1/3 to 1/4 of motor slip.

Flux Vector controls motor speed with measured shaft speed usually thru the use of a shaft-mounted encoder. Speed error is near zero in this system since the encoder will typically have 1024 pulses per revolution and the drive counts each one to determine speed and speed error.

 

[lukin1977]

Thanks DickDV

in the sensorless vector control what do you mean when you say: "at the time of commissioning"?

my english is not very good, sorry

lukin1977

 

[DickDV]

sorry, lukin1977, commissioning is when you first program the drive and adjust it to the application.

To build the motor model, on most drives you enter the motor nameplate data into the parameters and then start the drive and motor. The drive spends a minute or two accellerating and decellerating the motor and making a lot of measurements and then stops the motor. This is a sign that the model has been built properly. Some drives call this a "motor tune" but the function is similar. Note on these that if you change the motor or rewind it, you need to re-do the motor model so it matches the motor exactly.

 

[rockman7892]

Anybody know where to find a more detailed explanation or theory on sensorless vector control. I have tried on several attempts to find someting worthwhile but was never able to find anything good.

 

[lukin1977]

thanks DickDV

100% understood

 

[jraef]

Your answer was here all along by the way...
faq237-1062


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[rockman7892]

Jraef

Thanks for the link, it was very helpful and provided a good overview.

The one thing with sensorless vector that I am having a hard time understanding is how the drive is capable of supplying full torque at zero speed.

I understand the concept of vecotrs and how the sensorless vector control breaks the current into its magnetizing and torque producing vectors, but cannot grasp how this can be used to produce full torque at zero speed.

I assume it has something to do with supplying the right amount of flux to the motor at zero speed in order to supply the given torque, but wanted to learn more in depth how this was possible.

 

[itsmoked]

rockman7892; It is my understanding that the feedback used on sensorless drives is practically none existent near zero so they cannot do a good job at zero. So they would normally not be used in an application where you'd want accurate control at or near 0 RPM. Like a crane app. You would want to use a shaft sensor then.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[rockman7892]

itsmoked

Thanks I understand what you are getting at.

So setting asided zero speed for the sake of this case I guess I would be refereing to more or less full torque at "very low speeds"

As you mentioned, how could even a shaft encoder be used to provide full torque at zero speed using a flux vector control method?

 

[itsmoked]

Keep in mind full torque at a low speed isn't much power because T x speed => power.

I'm not 100% confident, but probably something around FLA and at the low speed's, greatly reduced voltage,(to keep V/Hz happy), will get you "full torque". This isn't that amazing. I'd be more amazed at full horsepower at near zero but then the motor's shaft would probably be pretty amazed too,(briefly).

Keith Cress
kcress -

http://www.flaminsystems.com

 

[cswilson]

rockman:

Keep in mind that, electromagnetically, the rotor only sees the slip frequency and the flux, not the mechanical speed. So its behavior at 60 Hz stator frequency and 58 Hz rotor frequency (1740 rpm for a 4-pole motor)with "full voltage" can be matched at 2 Hz stator frequency and 0 Hz (zero speed) rotor frequency with a voltage scaled down to produce the same flux.

With a shaft encoder, detecting the rotor speed is easy all the way down to zero speed. "Sensorless vector" drives basically use the motor model and drive voltage and current measurements to back out the EMF. As you get close to zero speed, there is very little "signal" left amidst the inevitable measurement noise, and this kind of falls apart.

 

[Marke]

The actual performance varies quite a bit between different drives, some do not really get full torque capacity until they are approaching 1 Hz, others work well below 5 Hz, but the difficulty is the model is based on "historical" measurements, (what has happened) and it take time for the full cycle to pass so the measurement can be made.

Best regards,

Mark Empson
L M Photonics Ltd

 

[rockman7892]

So it sounds like at these low speeds near 0 hz the drive is adjusting itself to produce the correct amount of flux in order to produce the required torque.

In other words it senses the torque of the motor based either upon a model or encoder feedback, and then varies the voltage and/or frequency to produce the required amount of flux.

I understand that at low freq there may be full torque but not much power because of the low speed, but will this torque still be enough to drive a given load. I understand that power is a direct product of speed and torque, but when does power come into play in regards to spinning a load. As long as the correct torques is supplied the load should spin regardless of speed or power? Is this correct?

 

[Marke]

Yes, a closed loop system uses an encoder to tel the VFD brain where the rotor is and s it is able to accurately control the slip torque etc.
The open loop vector system is reliant on a mathematical model to estimate the slip and so it becomes less effective at speeds close to zero due to insufficient information in a reasonable time.
If you need rated torque at zero shaft speed or very low frequencies, i.e. where you need a high breakaway torque such as starting a loaded conveyor, then use closed loop vector or DTC controlled drives. Some open loop vector systems give better results than others below 5 Hz so you can sometimes get away with an open loop vector in these applications.

As long as the correct torques is supplied the load should spin regardless of speed or power? Is this correct?

Provided that the torque developed by the motor is greater than or equal to the torque required to spin the load at that speed, the load will spin.

Best regards

Mark Empson
L M Photonics Ltd

 

[DickDV]

It's probably more understandable to leave power (hp or kw) out of the analysis completely. Just figure that a motor makes torque at a particular speed, including zero speed in some cases. Power is simply a calculated value based upon those two numbers (torque and speed).

 

[rockman7892]

Thanks for the information guys. I now understand from an overall standpoint how these control modes work.

Is there any examples avaliable anywhere to show a drive producing full or near full torque at low speeds. I was looking for an application example or something with numbers and drive output alogrithim to show exactly how the drive is controling the flux is a motor.

 

[jraef]

rockman7892,
If you have a VFD controlled hoist motor anywhere, that is the perfect example of a motor needing to develop full torque at zero speed. A hoist motor has a mechanical brake that prevents the load from dropping right? When you want to raise or lower the hoist, you must first release the brake. When you do, if the motor were not already producing full torque, the load would start dropping and spinning the motor backwards. Once the load began moving, by the time an open loop drive calculated the proper algorithm the added torque requirement to now overcome the back-spinning motor against that moving load would likely exceed what the motor was capable of and the load would drop. So hoist drives use what is called "torque proving" to test the drive output and motor response at zero speed to make sure it is putting out maximum torque before releasing the brake. If you can look a one of those drives and see the programming, you can sort of see what is going on, but as to the actual algorithm, all manufacturers I know of consider that to be IP.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[rockman7892]

Jraef

Your example is a good one. Thank you.


I am trying to quantify my understanding of the subject by looking at the relationship between Voltage & frequency, and flux, and therfore torque. I was looking for equations that show the amount of flux produced for a given voltage and frequency but have been unable to find any. I'm assuming that using the equivelent circuit model the more voltage the more stator current and therefore more fux. Also the less the frequency the long the voltage maintians the flux. Therefore the more flux, the more voltage induced in the rotor and therefore more rotor current leading to larger torque. Therefore keeping a constant V/Hz ratio keeps a constant flux and therefore a constant torque.

So saying all this I understand the contorl methods to works as follows:

For a constant V/Hz control method the drive varies the output voltage in proportion to the output speed. For a 480V drive the max output voltage is 480V at a max speed of 60hZ giving a V/Hz ratio of 8. For any speed below 60HZ the drive decreases the output voltage to maintain the V/Hz ratio of 8 and therefore maintain the same flux and therefore torque. The only problem with this control method is that you are assuming that the flux is staying the same based on the voltage adjustments but have no real feedback or indication of the exact flux or torque. There could be physical properties in the motor that do not allow it to produce the same exact flux at lower frequenciesa and this contorl method has no way of compensating.

For the sensorless vector, or flux contorl method the drive is constantly getting a torque feedback from either an encoder or based off a circuit model and alogrithim. It uses this torque feedback to monitor a constant torque. It then compares this torque at a certain frequency to mak sure that it is producing the required amount of torque. If for some reason the correct amount of torque is not being produced it can adjut the drive output voltage accordingly in order get the correct amount of torque. I would think that if the correct amount of torqe was not being supplied at a given speed, then the drive would slightly increase the ouput voltage thus increasing the V/Hz in order to get the correct torque. I would also guess that there were limitation set in place so that not to allow the motor to go into saturation with and increased V/Hz ratio.

So with the example of my 480V motor above, I would venture to guess that in order to produce full torque at near zero speed lets say 1 HZ then the voltage output would be held around 8V and if that wasn't enought the the dirve would increase this value until it saw the correct torque feedback.

This is my understanding of these contro methods from what I gained from this thread. Does it sound correct?

 

[lukin1977]

hi,

anyone knows which type of control is used for a Wire Drawing Machine application?

In my case the AC motors have a separate fan with inlet air filter for continous cooling. I know this probably doesn´t matter but I just mentioned anyway

Thanks,

lukin1977