VFD Help

Status
Not open for further replies.
I am looking into using a VFD for the following application:

A motor is attached to a shaft. On the other end, a constant torque is applied to the shaft, thus turning it at a constant, default speed. Tight control of shaft speed is needed both above and below this default speed.

In order to increase the shaft RPM above the default level, I was planning to use a VFD. The question then is: can I reduce the shaft RPM below the default level using the same VFD? If so, how would I do this, what type of VFD would I need, etc?

I’ve looked at DTC drives, would this be a good option? I’ve also read about braking with VFDs, but I couldn’t determine if there was a way to set the level of braking I wanted, or if it would just brake until 0 RPM?

Other info: Looking at 10HP motor @ 3600 RPM. Fine with just “wasting” energy in order to reduce shaft speed.

This is definitely all new to me, so I greatly appreciate any and all help.

Thanks.
 
VFD

VFD

Based on what you've described you need a closed loop system.....so you'll probably need a Flux Vector based inverter and a motor with an encoder. You can try with only a Flux Vector inverter open loop, as these will give you better speed and torque control over a standard inverter, but you haven't specified the speed tolerance nor what the application is.
 
VFD

VFD

The constant torque will spin the shaft at ~2,000 RPM (once the torque is turned on via a switch). I'd then like to use a VFD to increase the speed to ~3,000 RPM or reduce down to ~1,000 RPM with a tolerance of +/- 50 RPM.

Just did some preliminary research on closed vs open-loop Flux Vector inverters and looks like an open-loop one is probably sufficient for my needs?

Thanks.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
In a basic VFD, negative torque from regeneration is dealt with by using what's called "dynamic braking". The drive will absorb excess kinetic energy from the load as electrical power from the motor acting as a generator, store it temporarily on the DC bus and then fire a transistor into a resistor bank to dissipate that excess energy by burning it off as heat. The issue is with this is that in a typical "braking" operation, the duty cycle is expected to be limited and intermittent in that you use it to STOP the load and then sit for a while to cool off before the drive and DB components are being asked to do it again. In your case, it's conceivable that this could be continuous and generally Dynamic Braking is not a good choice for continuous operations. It could be done, but you would need to know a LOT about exactly how much energy you would need to dissipate and for how long, then select all of the components, INCLUDING the VFD, to be able to handle the power levels and duty cycle involved. This is not something for a novice to tackle in my opinion.

Another alternative is what's called a "Line Regenerative" VFD, one that pumps that excess energy back into the line source rather than burning it off into a resistor. In essence, the Line Regen VFD is more like having two back-to-back VFDs, so it will cost about the same. that would be a better option, but Line Regen capable VFDs at 10HP are not very common.

A third alternative, especially for a 10HP load, is to use a basic (Vector Control) VFD and add a stand-alone separate Line Regen module to it. There are "third party" companies like Bonitron that make these modules. Here is an example:
http://www.bonitron.com/m3645p.html
Now you'll notice that even they don't make one smaller than 36HP, but at least you don't need to buy a much bigger or more expensive VFD.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
The constant torque will spin the shaft at ~2,000 RPM (once the torque is turned on via a switch). I'd then like to use a VFD to increase the speed to ~3,000 RPM or reduce down to ~1,000 RPM with a tolerance of +/- 50 RPM.

Just did some preliminary research on closed vs open-loop Flux Vector inverters and looks like an open-loop one is probably sufficient for my needs?

Thanks.

probably. many VFDs these days have what is called sensor less vector mode that will keep you well within +/-50 RPM between 1000 and 3000 RPM. even pure open loop probably will.

as the previous poster mentioned, you will need some way to get rid of the excess energy, either via a brake resistor or line regen. if you go with a brake resistor make sure you buy one that can take 100% duty cycle at whatever load is required. most times brake resistors are rated with duty cycles as low as 5%. you can improve the duty cycle by forced air ventilation among other options, but you really need to make sure the resistor can take the heat continuously somehow.
 

Saturn_Europa

Senior Member
Location
Fishing Industry
Occupation
Electrician Limited License NC
Based on what you've described you need a closed loop system.....so you'll probably need a Flux Vector based inverter and a motor with an encoder. You can try with only a Flux Vector inverter open loop, as these will give you better speed and torque control over a standard inverter, but you haven't specified the speed tolerance nor what the application is.


I've only used encoders to measure position angle. Why wouldn't a tach be a better application for measuring 2,000 rpm?
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
I've only used encoders to measure position angle. Why wouldn't a tach be a better application for measuring 2,000 rpm?

encoders are more precise. and if you need to run in vector mode, you generally will need an encoder anyway.

you would be surprised how tight you can make a motor run with a tach on it in vector mode. about 20 years ago I did some machines that were running at about 1% of base speed with no cogging. I had the encoder also going into a PLC high speed counter and it was showing the thing dead nuts on at 1% of base speed. I was pretty impressed.
 
In a basic VFD, negative torque from regeneration is dealt with by using what's called "dynamic braking". The drive will absorb excess kinetic energy from the load as electrical power from the motor acting as a generator, store it temporarily on the DC bus and then fire a transistor into a resistor bank to dissipate that excess energy by burning it off as heat. The issue is with this is that in a typical "braking" operation, the duty cycle is expected to be limited and intermittent in that you use it to STOP the load and then sit for a while to cool off before the drive and DB components are being asked to do it again. In your case, it's conceivable that this could be continuous and generally Dynamic Braking is not a good choice for continuous operations. It could be done, but you would need to know a LOT about exactly how much energy you would need to dissipate and for how long, then select all of the components, INCLUDING the VFD, to be able to handle the power levels and duty cycle involved. This is not something for a novice to tackle in my opinion.

Another alternative is what's called a "Line Regenerative" VFD, one that pumps that excess energy back into the line source rather than burning it off into a resistor. In essence, the Line Regen VFD is more like having two back-to-back VFDs, so it will cost about the same. that would be a better option, but Line Regen capable VFDs at 10HP are not very common.

A third alternative, especially for a 10HP load, is to use a basic (Vector Control) VFD and add a stand-alone separate Line Regen module to it. There are "third party" companies like Bonitron that make these modules. Here is an example:
http://www.bonitron.com/m3645p.html
Now you'll notice that even they don't make one smaller than 36HP, but at least you don't need to buy a much bigger or more expensive VFD.

Thanks for all the info, Jraef, and for putting it in an easy to understand manner. It all makes sense...so now just have to figure out the best approach. You are correct in that we may run for several hours at 1,000 RPM, so sounds like either Option 2 or 3 is the best way to go.

For Option 2, would something like this work: http://www.industry.usa.siemens.com...120-vector-drive.aspx#Technical_20Information My guess is that yes it would work, but is probably very expensive and overkill for this application.

Option 3 sounds like probably the most economical way to go. I'll talk to Bonitron and go from there. Any recommendations of what make/model of VFD to combine with the Bonitron module?

Any other issues/considerations that you see with this overall setup?

Thanks again.
 

Fulthrotl

~Autocorrect is My Worst Enema.~
encoders are more precise. and if you need to run in vector mode, you generally will need an encoder anyway.

you would be surprised how tight you can make a motor run with a tach on it in vector mode. about 20 years ago I did some machines that were running at about 1% of base speed with no cogging. I had the encoder also going into a PLC high speed counter and it was showing the thing dead nuts on at 1% of base speed. I was pretty impressed.

that's what came to mind for me... an encoder to a plc counter, with a couple lines of code
to drive the vfd.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
Thanks for all the info, Jraef, and for putting it in an easy to understand manner. It all makes sense...so now just have to figure out the best approach. You are correct in that we may run for several hours at 1,000 RPM, so sounds like either Option 2 or 3 is the best way to go.

For Option 2, would something like this work: http://www.industry.usa.siemens.com...120-vector-drive.aspx#Technical_20Information My guess is that yes it would work, but is probably very expensive and overkill for this application.

Option 3 sounds like probably the most economical way to go. I'll talk to Bonitron and go from there. Any recommendations of what make/model of VFD to combine with the Bonitron module?

Any other issues/considerations that you see with this overall setup?

Thanks again.
I used to work for Siemens, that G120 drive is technically capable of line regeneration to a certain extent, so yes, it would probably work for your application. The thing is, it is not like other Line Regen drives, it is a somewhat unique design called a 'Fundamental Front End" (FFE) that is kind of a half-baked hybrid (for lack of a better term) way of doing it. Most other VFD mfrs who started down the road of Line Regen drives investigated FFE and rejected it because it causes more problems than it solves. Why Siemens chose to do it? It's cheaper and I think it's because they are so big all over the world, they can essentially force it down the throats of their big customers. Then since nobody else uses it, those customers will be stuck with Siemens.

In my opinion though, if you are not already VERY adept at using VFDs and navigating complex instruction manuals, Siemens is not the way to go. That drive is one of the most complicated drives I have ever come across in my 30 years in the drives business. In true German fashion it is a highly capable product and Siemens does not make junk, but oh boy, is it tough to use even for seasoned veterans. As we used to joke when I worked there, (using your best Sgt. Schultz impersonation) "There is the right vay, there is the wrong vay and there is the Siemens vay to do something, which is neither right nor wrong, but you VILL do it this vay."

Being that I get a paycheck from a VFD mfr and I am in your neck of the woods, it wouldn't be right for me to recommend a brand and direct you my way. But my suggestion is to stick to the major brands that will be supporting their products for the long run, for the most part everyone who's been in the business for any length of time HAS to make a good product, or they can't survive. The rest of the market goes to people who make CHEAP products, but they tend to change designs like they change underwear, or disappear altogether (as many of the super cheap Chinese products are doing now). Then find a LOCAL supplier you trust and who has a drive specialist on staff to help you. What you are going to do it totally possible, but not simple for a beginner, so you will need some hand-holding as you get going. 800 numbers are not the best way to go about that.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
that would be nice..... to encode, or not to encode, that is the question....

For a good SVC (Sensorless Vector Control) drive vs an FVC (Flux Vector Control) drive, SVC speed regulation will be +-0.4%, FVC with an encoder feedback will be +-0.001%. So better yes, but when it comes to velocity control for most machines, you are splitting hairs a little...

Where the encoder feedback option shines is when it comes to torque control. The encoder feedback lets the drive do a true vector control algorithm that allows the complete separation of the torque producing current from the magnetic flux current so that the drive can correct for torque requirements with less overshoot by keeping flux at only what is necessary. SVC cannot quite get there. Torque is BETTER with SVC than without it, but torque CONTROL as the primary issue is not where SVC shines and you need to go with full Flux Vector Control. Some mfrs using the latest generation of current sensors in the drive can now do FVC without encoders too, but those will not be the typical low cost drives people like to use in this HP range.

Even then, "Encoderless FVC" is not something I personally would trust on a hoist control package when you release the brakes on the hoists holding up a 747 (as I have done at Boeing). I want that "Full Torque at Zero Speed" that you need to do Torque Proving in that application. Torque proving means the drive knows how much torque a motor needs to apply BEFORE allowing the brake to release, so you have to be capable of creating that torque without having the shaft (and magnetic fields) moving first, which means knowing EXACTLY where that rotor shaft is at that moment with relation to the stator poles. With an incremental encoder, you are never off by more than 1/400th of an inch (depending on the encoder resolution).
 
Hello again,

I’m getting close to purchasing a line regen VFD for the application previously discussed in this thread. As such, I’m trying to understand the details better. I’ve been doing some research, but still have the following questions:

1) How do you wire up a line regen VFD – specifically, how does the regenerated energy return to the utility supply/grid? I know you bring in power and land it on the input terminals, but does the regenerated energy return to the grid through the same wiring, or is there a different set of output terminals on the drive? If there’s a different set of terminals/pathway for the regenerated energy, is some type of conditioner required to ensure the correct voltage and frequency returns to the grid? What about a way to meter it? Is coordination with the utility provider required? We’re not wanting to receive credit for this regenerated energy, just to get rid of it.

2) How will we control this application – specifically, will the VFD automatically determine whether energy should flow to or away from the motor based on the process dynamics? For example: let’s assume we turn on the VFD and motor and send a 12 mA signal to the VFD, which should correspond to a VFD output of 30 Hz and thus a shaft speed of ~1800 RPM (assuming a 2 pole motor). Then, we introduce the external force to the shaft. This force is strong enough to turn the shaft at 2500 RPM on its own. So, in this situation, will the VFD automatically send enough energy back to the grid to continuously brake the shaft down to 1800 RPM? Are any other inputs (such as feedback about the shaft speed) required to properly operate in this manner?

As I mentioned previously, this is all new to me, so feel free to correct any incorrect terminology or errors in my questions. This is definitely fun, interesting, and very practical to learn, so as always, any help is greatly appreciated.

Thanks.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
Hello again,

I’m getting close to purchasing a line regen VFD for the application previously discussed in this thread. As such, I’m trying to understand the details better. I’ve been doing some research, but still have the following questions:

1) How do you wire up a line regen VFD – specifically, how does the regenerated energy return to the utility supply/grid? I know you bring in power and land it on the input terminals, but does the regenerated energy return to the grid through the same wiring, or is there a different set of output terminals on the drive? If there’s a different set of terminals/pathway for the regenerated energy, is some type of conditioner required to ensure the correct voltage and frequency returns to the grid? What about a way to meter it? Is coordination with the utility provider required? We’re not wanting to receive credit for this regenerated energy, just to get rid of it.

2) How will we control this application – specifically, will the VFD automatically determine whether energy should flow to or away from the motor based on the process dynamics? For example: let’s assume we turn on the VFD and motor and send a 12 mA signal to the VFD, which should correspond to a VFD output of 30 Hz and thus a shaft speed of ~1800 RPM (assuming a 2 pole motor). Then, we introduce the external force to the shaft. This force is strong enough to turn the shaft at 2500 RPM on its own. So, in this situation, will the VFD automatically send enough energy back to the grid to continuously brake the shaft down to 1800 RPM? Are any other inputs (such as feedback about the shaft speed) required to properly operate in this manner?

As I mentioned previously, this is all new to me, so feel free to correct any incorrect terminology or errors in my questions. This is definitely fun, interesting, and very practical to learn, so as always, any help is greatly appreciated.

Thanks.
The ones I have seen use the same power input terminals to feedback power on the line, but my sample size is pretty small.

From the user perspective it is pretty much transparent as long as the parameters are set correctly. If there is energy flowing into the load side of the drive it will direct it back to the power source.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
Yes, same terminals, no differences in wiring.

The vector control drive just does what you tell it to do in terms of the speed command. So if you want it running at 30Hz, that's what it does. A vector drive is different because it is actively tracking the real speed of the motor, rather than just hoping it stays the same. If in the process of maintaining that 30Hz output speed the load attempts to go faster, the drive will automatically compensate the output to maintain that speed by lowering the output frequency given to the motor accordingly, for example 27Hz. When the motor shaft speed is higher than the frequency given to it, the motor becomes a generator and the excess energy goes into the DC bus. From there, the active front end drive then fires the converter devices so that the excess power can flow back to the utility.
 
Status
Not open for further replies.
Top