VFD Modes of operation

[steve66]

Can anyone provide a little background or anything to explain this sentence taken from a VFD Cutsheet:

Offering three modes of motor control: V/Hz, Sensorless Vector and Flux Vector, the ACS550 performs accurate speed and torque control of any standard squirrel cage motor.

What's the difference between V/HZ, Sensorless Vector, and FLux Vector?

And whats the difference between speed and torque control?? Don't we give all VFD's a speed setpoint?

Finally, exactly what is the difference between a VFD set for "variable torque" and one for constant torque? Another manufacturer provides these options.

Steve

 

[steve66]

Nobody has any input?

Come on...don't make me ask the sales rep

 

[Jraef]

Can anyone provide a little background or anything to explain this sentence taken from a VFD Cutsheet:



What's the difference between V/HZ, Sensorless Vector, and FLux Vector?

And whats the difference between speed and torque control?? Don't we give all VFD's a speed setpoint?

Finally, exactly what is the difference between a VFD set for "variable torque" and one for constant torque? Another manufacturer provides these options.

Steve

V/HZ, Sensorless Vector, and FLux Vector:
V/Hz = VFD does just that, changes V and Hz to change speed. Has no idea if that accomplished what you wanted, no feedback.
Vector (in general): uses feedback signals to determine what happened in the motor, then adjusts the vector of V/Hz to maximize the output to match your criteria; i.e. speed or torque etc. Vector control is divided by the type of feedback:
Sensorless Vector = the feedback to the drive's mP is taken from the current signal of the power itself as it goes to the motor. the VFD creates and remembers a mathematical model of the motor performance and as it watches that performance, uses the error it sees to tweak the output.
Flux Vector = the feedback is much more accurate because it comes from an encoder attached to the motor shaft. So like a servo, the mP is watching exactly where the rotor shaft is at any given moment so that it can tweak that output pattern to maximize the flux in the motor and provide high (highest) precision in the output performance. With this you can achieve the holy grail of AC motor use: replacing a DC drive and motor so that you can get 100% torque at zero speed, i.e. a hoist motor that needs to make absolute sure the motor is in control of the load before a brake releases it. Also, just like a servo, you can do rudimentary positioning control. Servos are better at it, but try to find a 200HP servo!

And whats the difference between speed and torque control?? Don't we give all VFD's a speed setpoint?

Maybe yes, maybe no. Ever watched a wire winder? You have overall speed control, but the pay-out and take-up reels are rapidly changing speeds relative to each other and if you don't precisely control torque, will stretch the wire. So for winders, you typically use a Torque setpoint. All this means is, the VFD is smart enough to be set to do either. A V/Hz only drive (although now rare) will not be able to do this.

Finally, exactly what is the difference between a VFD set for "variable torque" and one for constant torque? Another manufacturer provides these options.

Any drive is capable if controlling either, this is mainly a marketing issue. A VT load, like most centrifugal pumps for instance, will never need full torque at reduced speed, because the nature of the load changes quadratically with speed. that means that the HP reduces at the cube of the speed reduction; at 50% speed for example, the LOAD will draw .5 x .5 x .5 of the HP it would at full speed, so that's 12.5% HP, which means the VFD output current requirement will vary as well. So because it's accepted that the VFD will probably never get over loaded, they can play a "numbers game" with the ratings. So what would work for a 50HP Constant Torque load, may be OK for a 60HP VT load and therefore will be a little cheaper for that 60HP VT application. That's all it means.

 

[rcwilson]

Steve- I've been burned by this before and found some good google searches. EC&M had a series of articles that did a good job of explaining this. I'm not a VFD expert but I'll try to give you answers at the risk of giving you the wrong ones.

Modern VFD's can use software to compute speed of the motor based on anaylzing the motor current. This calcuated speed can be used to control load speed pretty close without using a tach feedback, but with some speed error. This is the Sensorless Vector Drive. The "vector" terminology comes from the analysis that is used to resolve the current measurement into components for magnetizing current and load torque and comparing them to the voltage waveform. Program 1800 rpm into this drive on an 4-pole (1800 rpm) motor and it will put out 62 Hz or whatever frequency is needed to hold 1800 rpm instead of the 1780 rpm on the motor nameplate.

For more precise speed and position control ,add a tach or a resolver to the load shaft and provide feedback to the VFD. The Flux Vector Drives use similar calculations to precisely monitor and control the torque in the motor to keep the tachometer or resolver speed and position at setpoint.

The V/Hz drive doesn't worry about feedback or load speed. It just puts out the frequency and lets the motor run at the speed it wants. That 1800 rpm motor will run at 1780 rpm and the drive won't care. This mode is ideal for pumps and fans or anything where precise speed isn't important, you just need it to go faster or slower. The V/Hz control keeps the volts to frequency ratio constant to avoid overfluxing (saturating) the motor iron. At 50% speed max voltage should be 50% of nameplate to keep constant V/Hz.

If you put a sensorless flux drive on a cooling tower fan you get great control that you don't need. To control torque, the drive appears to reduce voltage so the current is higher to maintain the same HP output. When a gust of wind speeds up the fan, the drive senses the speed change and supplies negative torque to bring the speed back. As a result, the fan current is always spiking as the drive tries to maintian that 900 rpm speed +/-2 rpm, when all you needed was something more than 850 and less than 1000 rpm. A V/Hz control mode would be a lot better, easier to tune and less strenuous on the motor. (I won't go into how I learned this.)

Fans and centrifugal pumps are variable torque loads. At 50% speed the torque is 50% so the HP needed is only 25%. A conveyor takes the same torque whether it is running 1 foot per minute or 200 fpm, at 50% speed it stillneeds ful torque adn 50% HP. A drive programed for variable torque loads can do some things at reduced speeds to minimize losses that a constant torque drive can't wihtout stalling the load.

 

[rcwilson]

Delete my musings. I bow to Jraef who has helped me a lot on VFD's. He posted when I was meandering.

 

[Jraef]

Delete my musings. I bow to Jraef who has helped me a lot on VFD's. He posted when I was meandering.

I like your cooling tower example, that's one I have never come across.

 

[John120/240]

Jraef, thank you for explaining the topic in easy to understand lanuage.
Who is your avatar? Star Trek?
John

 

[Besoeker]

Ever watched a wire winder? You have overall speed control, but the pay-out and take-up reels are rapidly changing speeds relative to each other and if you don't precisely control torque, will stretch the wire. So for winders, you typically use a Torque setpoint.

If they are changing speed relative to each other because of changing unwind and rewind diameters, what you need is constant power to keep constant tension in the wire.
Power is speed times tension. For constant linear wire speed, you thus need constant power.

 

[jcole]

I dont mean to hack thread. Just a few questions to add.

So which mode should you use for a variable torque load? constant torque load? I think your going to say it doesnt matter as long as what ever mode I choose provides enough starting torque and speed control that I need. Is that correct?

How do you program a drive for variable torque or constant torque? I never seen one that had a parameter to select either? Is it done by which mode I select?

We had new pumps installed that the manufactor said had to have constant torque drives to control motor. The old pumps were centrifucal type and had a powerflex 400 fan/pump (v/hz) drive. The new pumps were progressive cavity (positive displacement type).

I tired to convince my supervisor not to buy new drives because our old drives would probaly provide enough starting torque for our application (low pressure, high volume sludge pumping). He went on and spent the $8000 needed to get the powerflex 70 drives, which provide constant torque control.

I set the mode to v/hz and the pumps have ran fine for over a year now.

Why would the manufactor recommend us change the drives? Would you have changed the drives? Would you set it up on vector control?

Thanks for replies.

 

[steve66]

Good answers. Thanks a million.

I dont mean to hack thread. Just a few questions to add.

Feel free to hack (or hijack) away. Those questions fit right in.

Steve

 

[Jraef]

I dont mean to hack thread. Just a few questions to add.

So which mode should you use for a variable torque load? constant torque load? I think your going to say it doesnt matter as long as what ever mode I choose provides enough starting torque and speed control that I need. Is that correct?

How do you program a drive for variable torque or constant torque? I never seen one that had a parameter to select either? Is it done by which mode I select?

We had new pumps installed that the manufactor said had to have constant torque drives to control motor. The old pumps were centrifucal type and had a powerflex 400 fan/pump (v/hz) drive. The new pumps were progressive cavity (positive displacement type).

I tired to convince my supervisor not to buy new drives because our old drives would probaly provide enough starting torque for our application (low pressure, high volume sludge pumping). He went on and spent the $8000 needed to get the powerflex 70 drives, which provide constant torque control.

I set the mode to v/hz and the pumps have ran fine for over a year now.

Why would the manufactor recommend us change the drives? Would you have changed the drives? Would you set it up on vector control?

Thanks for replies.

Answered in order of my random though process...

Progressive cavity pumps are CT loads, you need torque to move that wiggle stick consistently throughout the speed range, the load (in terms of torque requirements) never changes.

As I said earlier, VT vs CT ratings is just a numbers game. If your HP was the same after the pump change, then the VT rated drive may have been too small and you would have had nuisance tripping. But if the PD pump HP was less than the old centrifugal, a change in drives may have been an exercise in budget inflation. You didn't say.

Most VT loads will work fine with V/Hz mode and it's a lot easier to set up, so no need to make an application more complicated by using vector control unnecessarily. Most CT loads will benefit from having better torque / overload capabilities by using vector control. A PD pump is kind of a toss up, but I would use sensorless vector for a sewage sludge pump, you never know what's coming down the pipe...

Why did the mfr recommend new drives? A) they knew more about the application than you have posted here, or B) they smelled an opportunity to weasel $8K out of someone ignorant of the issues. Only you can determine which one it was.

How to you program the difference? Usually in the maximum short time OL current settings. If the VFD is strictly a VT drive (some mfrs do this), then it will not give you the option of setting the drive (not motor) OL point to 150% for 1 minute, it will limit you to 110% or 120% for 30 seconds (varies by mfrs). You can also take extra advantage of a load being VT and program what is called an "energy saver" algorithm into the drive that lowers the voltage a little more with speed drop in order to save on some of the magnetizing current for the motor when you don't need as much power. Just to be clear though, most CT rated drives can be "turned down" to VT, but not necessarily vice versa. UL has issues here in the US with the nameplate ratings, that's why some mfrs choose to limit the adjustments on their strictly VT rated versions; it creates no conflict regarding what the VFD is capable of and how the label reads.

 

[Jraef]

If they are changing speed relative to each other because of changing unwind and rewind diameters, what you need is constant power to keep constant tension in the wire.
Power is speed times tension. For constant linear wire speed, you thus need constant power.

You and I know that, I was trying to keep it simple. Most advanced drives now have in excess of 200 programmable functions, it takes a lot of training and experience to tackle something like a wire winder. Most electricians who get involved with VFDs will never need to use torque control, let alone torque control with a speed follower, even once in their careers. I was using it as an over simplified example to respond to the "why would you do that?" question. But thanks for clarifying that, I know there are more advanced users who use this forum.

 

[philly]

Modern VFD's can use software to compute speed of the motor based on anaylzing the motor current. This calcuated speed can be used to control load speed pretty close without using a tach feedback, but with some speed error. This is the Sensorless Vector Drive. The "vector" terminology comes from the analysis that is used to resolve the current measurement into components for magnetizing current and load torque and comparing them to the voltage waveform.

I've always wondered exactly how this works. Can someone explain it using the motor equivelent circuit model and its equations?

I understand how the current can be resolved into it's two components but am not sure what the drive does with these compoents for calculations?

Can it use these values to determine the angle of the induced magnetic field in the rotor (Bsync vector) and determine the angle of the of the rotor field resulting from the rotor current. Then by adding these two vectors you can come up with a magnitude and angle for the Bnet field which is the rotor field that interacts with the induced field from the stator to produce the torque. By then comparing this Bnet vector to the Bsync vector can the angle difference be determined to calculate a slip and therefore speed? Am I on the right track with this?

 

[Besoeker]

You and I know that, I was trying to keep it simple.

Keeping it simple should not detract from accuracy.