Hello all!

I am developing an application where a rope is wound around a reel. The rope is pulled out as far as 100 feet. A quadrature encoder is used to trace the speed of the rope during payout via a cam follower.

The reel is attached to a brushed DC Gearmotor. The application requires that the rope maintain tension the entire time. When the force pulling the rope ceases, the motor should wind the reel back up.

I cannot use a clutch due to complexity and cost for this application.
I cannot use a spring rewind due to the distance of payout.

My plan was to oversize the motor and run it at such a low current and voltage level, that it would be back-drivable, yet maintain tension.
This would also allow the motor to begin retrieving the rope once the payout force is removed.

I am looking for advice on the feasibly of this idea. As well as alternative solutions.

I really appreciate any help or advice you can provide.

Thank You!

-Chuck

 

It would appear that the motor never rotates under power on the payout, only tension and rewind is needed, if I read right?
You could try a dynamic brake by placing a very low resistor or a short across the armature when paying out, when the encoder senses a stop then the brake is removed and necessary power applied to the motor to reel in.
Max.

 

It would appear that the motor never rotates under power on the payout, only tension and rewind is needed, if I read right?
You could try a dynamic brake by placing a very low resistor or a short across the armature when paying out, when the encoder senses a stop then the brake is removed and necessary power applied to the motor to reel in.
Max.

Max,
The motor is rigidly fixed to the reel, it must rotate in the opposite direction of power (physically being back-driven) while payout is happening. Think of how a measuring tape re-wind functions. I am looking for this type of functionality, with a motor.

I really appreciate the suggestion!!

 

It would appear that the motor never rotates under power on the payout, only tension and rewind is needed, if I read right?
You could try a dynamic brake by placing a very low resistor or a short across the armature when paying out, when the encoder senses a stop then the brake is removed and necessary power applied to the motor to reel in.
Max.

Max,

After reading over your suggestion a few times I believe I didn't understand you correctly. it seems that your suggestion could work well! Could you explain in more detail what you are thinking?

Cheers!

-Chuck

 

Back driving a motor usually means it shaft or operator is rotated by another source, i.e. no power is applied, in this case it would be whatever is pulling the rope out.
So if the motor in this case were shorted or resistive braked it would apply tension that would increase as the pulling speed increased, the brake would be disconnected when it is required to reel in and motor power would be applied to reel the rope back in.
It all depends on all the mechanical constraints of the set up you have, as all we have so far is a brief description.
Max.

 

Hello there,

Looking briefly at this problem, it appears that it is possible that dynamic braking isnt enough, it has to be driven in reverse, but keep in mind that does not mean that it will be turning in reverse, as during normal operation it will still be turning in the forward (payout) direction. The reverse power will keep the tension regardless what the force going out is. Also keep in mind this does not mean that it will not be strickly speaking still performing dynamic braking, it may well do that on it's own, so we can kind of call this "dynamic braking squared", where we get brake action sometimes and actual reverse force other times, depending on the payout force at the time, and since the payout force may vary considerably, the reverse force from the motor may vary just as much.

It sounds like a feedback loop control type system. Measure the speed of the payout (with your tach) and adjust the reverse motor current accordingly until it gets close to the desired (i assume constant) speed. The actual implementation depends on what knd of control system you want to use.

 

Back driving a motor usually means it shaft or operator is rotated by another source, i.e. no power is applied, in this case it would be whatever is pulling the rope out.
So if the motor in this case were shorted or resistive braked it would apply tension that would increase as the pulling speed increased, the brake would be disconnected when it is required to reel in and motor power would be applied to reel the rope back in.
It all depends on all the mechanical constraints of the set up you have, as all we have so far is a brief description.
Max.

Max,

I really like your approach to this problem. I apologize for not covering all of the mechanical constraints in the first post.

The object that pulls the rope out may travel at speeds of 5 m/s away from the device. This object may stop its motion at anytime. It may just stay put where it is, or it may start traveling back towards the device. The device must maintain tension at all times, and must take up slack when the object begins approaching the device.

I am worried about the response time of the solution you are introducing. I am afraid it could generate slack.

Thank you So much!

-Chuck

 

Hello there,

Looking briefly at this problem, it appears that it is possible that dynamic braking isnt enough, it has to be driven in reverse, but keep in mind that does not mean that it will be turning in reverse, as during normal operation it will still be turning in the forward (payout) direction. The reverse power will keep the tension regardless what the force going out is. Also keep in mind this does not mean that it will not be strickly speaking still performing dynamic braking, it may well do that on it's own, so we can kind of call this "dynamic braking squared", where we get brake action sometimes and actual reverse force other times, depending on the payout force at the time, and since the payout force may vary considerably, the reverse force from the motor may vary just as much.

It sounds like a feedback loop control type system. Measure the speed of the payout (with your tach) and adjust the reverse motor current accordingly until it gets close to the desired (i assume constant) speed. The actual implementation depends on what knd of control system you want to use.

MrAl,

Thank you so much for your input.

The speed of the object pulling the rope may vary between 0 m/s and 5 m/s.
Our processor is a Nordic NRF52.

We plan to design our own circuit board for this device, but seeing as it is intended to be a consumer product, we are looking to keep costs to a minimum. With this information in mind, how would you go about designing such a system?

Cheers!
Chuck

 

One problem as I see it is the payout rate will vary and a constant tension is required.
If the idea is to retain constant tension, a simple tension sensor could be constructed from the often basic used idea of 3 rollers with the centre providing tension sensor and carry the measuring device, pot etc.
This would provide automatic control as to feed, tension and rewind.
Bi-directional motor control would be used in this application.
http://www.tensionmeters.com/design.htm
Max.

 

One problem as I see it is the payout rate will vary and a constant tension is required.
If the idea is to retain constant tension, a simple tension sensor could be constructed from the often basic used idea of 3 rollers with the centre providing tension sensor and carry the measuring device, pot etc.
This would provide automatic control as to feed, tension and rewind.
Bi-directional motor control would be used in this application.
http://www.tensionmeters.com/design.htm
Max.

Max,

The line just needs to stay taught so we can precisely track the speed of the object.
The object may travel forwards, backwards or stop.

Cheers! Thanks again!
-Chuck

 

What you want is one of these.

 

So some type of tension sensor may work to provide the motor direction FB needed.
From what you have indicated so far, I presume you only need to know the speed of the object not control the speed itself?
Max.

 

What you want is one of these.

I wish, but we need up to 25 meters.

 

So some type of tension sensor may work to provide the motor direction FB needed.
From what you have indicated so far, I presume you only need to know the speed of the object not control the speed itself?
Max.

That is correct, this system is for tracing the speed of the object only. Not putting any force on the object.
However, This type of tension system really increases manufacturing and production costs. We can do it, but I want to keep the design as cost effective and simple as possible.

 

Hello again,

Oh so you dont need to control the speed then just the tension?
That's a different story.

One way is to construct a spring loaded arm that pivots with a pulley at the free end. The 'rope' goes over the pulley and the load is somewhat perpendicular to the arm during normal operation. As the tension increases, the arm naturally pivots forward, and that tells you that the tension has increased. As the tension decreases, the arm pivots back toward rest position and that tells you that the tension has decreased. Keeping the arm in the same position would be the job of the motor, and the motor would be controlled from a measurement of the arm angle.
For this kind of mechanism which is just one example, you would work up the equations for the change in angle with tension and relate that information to the controller which then controls the current to the motor. If the controller for example integrates the error in the angle and uses that to control the motor, the steady state error in tension should be small.

This is starting to remind me of a fishing reel line spooler, that keeps tension on the line as the user reels up more and more line until the reel is full. That's for replacing the fishing line on a fishing reel with new line.
You might gain some insight if you could find something like that on the web.

 

I did an open-loop one way tension controller once using a brake.

A lever arm with roller attached to a potentiometer with a spring is pretty much a tension gage. So, a funky servo motor just has to decide which way to go and how fast it's allowed to do it.

 

For rope I envision 3/32 in. nylon line, with tension of about 6 oz. This might be handled with a continuous rotation servo. Need to check torque available. A spring loaded idler is set for 6 oz. at mid position & is attached to a direct reading shaft encoder for motor speed & direction ??
Mine is listed at 66.7 oz. in.

Sorry, 70 RPM will not cut it, 2 ft./ s vs. needed 16 ft./s. Need about 1000 RPM for 4 in dia. drum ?

 

I am developing an application where a rope is wound around a reel. The rope is pulled out as far as 100 feet. A quadrature encoder is used to trace the speed of the rope during payout via a cam follower.
The reel is attached to a brushed DC Gearmotor. The application requires that the rope maintain tension the entire time. When the force pulling the rope ceases, the motor should wind the reel back up.

Another observation, If this is a typical worm and pinion gear box, they cannot typically be back driven, if this is the case then the motor would have to be in feed mode when ever the rope is pulled, have you tested this in the one you have?
Max.

 

Another observation, If this is a typical worm and pinion gear box, they cannot typically be back driven, if this is the case then the motor would have to be in feed mode when ever the rope is pulled, have you tested this in the one you have?
Max.

Max,
Ive made sure to use a Spur Gear Box. and the very small gearing ratio makes back driving easier.

 

Max,
Ive made sure to use a Spur Gear Box. and the very small gearing ratio makes back driving easier.

I was thinking something like this...

The Hg switch is a mercury switch or any tilt switch that activates when it is not level (or rotate the switch to any angle that will kick it on when needed).