Adaptative motor torque

Learn the NEC with Mike Holt now!
Status
Not open for further replies.

SuperPat

Member
Location
Belgium
Occupation
programmer
Hi all,
Here is my first question on this forum.
I hope I'm not mistaken about the topic.
My english is not very goog .....
I would like to solve a synchronization problem between 2 motors.

This involves rolling up a paper reel (max 20Kg) from the left reel to the right reel.
The right reel is tractor and the left spool is freewheeling.
The speed must be constant and a maximum of 2 m / s
To calculate the speed, I have put an encoder between the 2 motors.
My problem is the tape during the winding must always remain taut (stretched).

I thought of stepper motors, but I will not be able to brake the unwinding reel.
I thought of synchronous motors for its constant speed despite the load, but I also don't know how to brake the left reel with these motors.
Then DC motors but I have to adapt the torque as the reel becomes heavier.

Do you have an idea to solve this problem?

Thank you

Patrick
 

LarryFine

Master Electrician Electric Contractor Richmond VA
Location
Henrico County, VA
Occupation
Electrical Contractor
Pat, welcome to the forum. Your English is perfect. (y)

If the right reel is driven at a constant RPM, the paper speed will increase as it fills. It seems you want to regulate the paper speed, which makes me think of a capstan-and-pressure-roller setup, similar to a reel-t-reel tape deck.

1588252788023.jpeg 1588252807993.png

Perhaps, you can find a way to drive the paper speed itself, and have the right reel merely maintain tension and roll up the paper.

Or, maybe you can just monitor the paper speed, and vary the right reel speed as necessary to maintain the desired paper speed.
 

PaulMmn

Senior Member
Location
Union, KY, USA
Occupation
EIT - Engineer in Training, Lafayette College
The supply reel will need to supply some drag/tension on the reel so that the paper has to be pulled off of the supply reel. My reel-to-reel tape recorder definitely did not let the supply reel free-wheel. Free-wheeling will be like a roll of toilet paper and either a cat or a small child-- One tug on the end of the paper, and you have piles of paper all over the floor!

I don't think you need a capstan/pinch roller, but you do need a tachometer/encoder wheel that presses against the paper and gives out pulses or other indication of how fast the paper is moving. This will then control the speed of the takeup motor.

Tension on the supply reel could be as simple as a drag wheel on the shaft of the reel so that it's not free-wheeling.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
What size motors are you thinking of here? This is a classic "torque control" application, for which good quality AC Flux Vector Drives or 4 quardrant DC drives are well suited. But if this is really small, servos may be a better choice, I don't quite understand your not wanting to use them. Maybe you could explain that.
 

RumRunner

Senior Member
Location
SCV Ca, USA
Occupation
Retired EE
Try this site:

Haven’t used it since I was actively involved in manufacturing years ago. I'm most familiar with magnetic coupling and air friction brakes.


This is not an endorsement of any product or the manufacturer. . . just might give you an idea of what's available out there.

  • succes, veel geluk!
Good Luck
English is great. You speak Dutch or French?
 

SuperPat

Member
Location
Belgium
Occupation
programmer
Thank you for all the answers.
I will use well on an encoder to be able to adjust the speed. Because the more the reel fills, the more I will have to slow down to keep my speed.
I believe that with DC motors would also be possible but more difficult for the adaptation of the torque.
What worries me is this situation:
When the reel that unwinds (at 2 m / s) reaches the end of the roll, how will it behave?
Won't it take place too quickly and relax the tension of the tape?
Should I brake it?
I am belgian (french)

regards
Patrick
 

paulengr

Senior Member
This is a classic problem and well known. I wouldn’t bother with steppers for the fact that it’s the wrong motor. A stepper has a very linear torque/speed curve so great for low end torque not to mention simplicity but the application calls for even torque over the entire speed range. This application is vastly better suited to PMDCs or BLDCs on the small end or AC induction motors on the larger end.

What you are doing is web tension control or winders.




https://www.bannerengineering.com/u...ction/roll-diameter-and-tension-control.htmlr.

What makes winders so deceptively simple is that it seems like if you can somehow synchronize/match the speeds just right everything will work well but it never, ever does. There are two trucks to it. The first is that one drive is a speed controller. The other is a torque controller. The speed controller can run constant speed but as you said the paper roll will go faster and faster usually until it rips. So you have to either control speed based on encoder feedback from a speed sensor on the unrolled paper and make that constant or control speed based on the diameter of the roll. Both methods work pretty well. This isn’t critical but if you don’t run constant speed paper (not roll) the speed demand on the other drive will run it beyond the motor speed range at some point.

The torque drive just runs at constant torque. The torque is chosen to keep enough tension on the paper so it spools up evenly on the roll. Too much and it breaks, too little and the paper gets loose on the roll. At first the drive will be going at maximum speed then it gradually slows down to very slow while your speed drive is doing the opposite,

Now here’s the interesting part, which drive is speed and which torque? The answer is it doesn’t matter but they both can’t be speed or both torque. And a second interesting fact...if the “unwinded” supplies torque it is 100% braking or regenerating. So if this is a large application you can use a common DC bus arrangement. In VFDs it doesn’t even need a front end and can run 100% disconnected because the inverter is always supplying power, not using it. So it will just feed power to the other (winding/speed) drive. Maximum power draw curiously is sort of in the middle because of this, not at the end points. And due to the fact that regeneration is not quite as efficient as generation (1st vs 2nd quadrant) you really want to do this or buy much bigger drives than you need.

This used to be an expensive thing to do. Today’s micro drives easily have torque and speed control built in.

As to encoders, it depends on how tight the torque regulation has to be. I’d try it with just sensorless control first if you are doing AC. On DC servos it comes with the servo in most cases and you kind of need it. You can use dancers and other primitive mechanical controls but that’s not modern torque control.
 

SuperPat

Member
Location
Belgium
Occupation
programmer
I must keep the tape stretched because it must pass through an ultrasonic cell (which counts the labels) which is very sensitive to vibrations.
And I would not like that when braking, the left roller continues to unwind.
I will use an encoder to adapt the speed.
The reel will be placed directly on the chassîs on "iglidur" strips.
A very slippery material.
So the driven roller will be automatically braked.
I prefer to avoid dancer.
Thank you
Patrick
 

bwat

EE
Location
NC
Occupation
EE
This is a classic problem and well known. I wouldn’t bother with steppers for the fact that it’s the wrong motor. A stepper has a very linear torque/speed curve so great for low end torque not to mention simplicity but the application calls for even torque over the entire speed range. This application is vastly better suited to PMDCs or BLDCs on the small end or AC induction motors on the larger end.

What you are doing is web tension control or winders.




https://www.bannerengineering.com/u...ction/roll-diameter-and-tension-control.htmlr.

What makes winders so deceptively simple is that it seems like if you can somehow synchronize/match the speeds just right everything will work well but it never, ever does. There are two trucks to it. The first is that one drive is a speed controller. The other is a torque controller. The speed controller can run constant speed but as you said the paper roll will go faster and faster usually until it rips. So you have to either control speed based on encoder feedback from a speed sensor on the unrolled paper and make that constant or control speed based on the diameter of the roll. Both methods work pretty well. This isn’t critical but if you don’t run constant speed paper (not roll) the speed demand on the other drive will run it beyond the motor speed range at some point.

The torque drive just runs at constant torque. The torque is chosen to keep enough tension on the paper so it spools up evenly on the roll. Too much and it breaks, too little and the paper gets loose on the roll. At first the drive will be going at maximum speed then it gradually slows down to very slow while your speed drive is doing the opposite,

Now here’s the interesting part, which drive is speed and which torque? The answer is it doesn’t matter but they both can’t be speed or both torque. And a second interesting fact...if the “unwinded” supplies torque it is 100% braking or regenerating. So if this is a large application you can use a common DC bus arrangement. In VFDs it doesn’t even need a front end and can run 100% disconnected because the inverter is always supplying power, not using it. So it will just feed power to the other (winding/speed) drive. Maximum power draw curiously is sort of in the middle because of this, not at the end points. And due to the fact that regeneration is not quite as efficient as generation (1st vs 2nd quadrant) you really want to do this or buy much bigger drives than you need.

This used to be an expensive thing to do. Today’s micro drives easily have torque and speed control built in.

As to encoders, it depends on how tight the torque regulation has to be. I’d try it with just sensorless control first if you are doing AC. On DC servos it comes with the servo in most cases and you kind of need it. You can use dancers and other primitive mechanical controls but that’s not modern torque control.
A ton of great info here, but isn’t it constant tension and not constant torque? If your rewind is doing the speed control to match a constant specified line speed, then your unwind would be wanting to keep a constant tension which would be a torque control scheme to achieve a desired constant tension. Essentially the torque set point needs to change as the roll changes diameter, no?

(Ignoring taper tension schemes)
 

SuperPat

Member
Location
Belgium
Occupation
programmer
Correct.
The tension of the tape must be "correct" but not necessarily constant (without breaking it :)).
So it seems to me that the torque motor must be adapted as the reel gets bigger
 

GeorgeB

ElectroHydraulics engineer (retired)
Location
Greenville SC
Occupation
Retired
When I did this sort of tension control thing in the low-tech world (textiles), we weighted (sometimes single pass, sometimes festoon accumulator design) with feedback on a shaft. In higher tech, as in plastic film, load cells were used. Intelligent controllers commanded the motors, the speed of the winder determined by an idler on the take-up.
 

paulengr

Senior Member
Torque is rotary force. If you are accelerating it’s positive (1st quadrant, motoring). If you are decelerating it’s negative. Mechanically drive/driven can be switched. As to slippery that’s controlled by roll sizes wrap angle, roll coatings (rubber vs Teflon) and tension. Multiple rolls if it’s really bad.

In an in house tire machine at Bridgestone they have an interesting let’s just say positive tensioner. The green rubber piece is really soft and tears easily. They let it sag in a big loop by gravity and have a distance sensor on the loop that controls the unwinder speed...sort of a positive torque loop. As the machine feeds material (speed control) it keeps the loop (tension) constant.
 
Status
Not open for further replies.
Top