Design problem with a constant torque motor

[pete]

TL;DR Summary

Run into a variation on the output of this constant torque spring assemble and am looking for advice on the design to remove this.

I’ve been trying to design a counterbalance system using a constant torque motor spring assembly. The spring I’m looking at has 25.4kg cm torque on the shaft from the output drum.

So if I put a 100mm diameter cable drum on this shaft I should be able to counterbalance 5kg over the given extension of the spring. I’m unclear on how to calculate the exact amount of revolutions of the output shaft but it should be in the ballpark of about 21. Let's say I use a 1mm diameter Dyneema cable. I can go smaller but I’d like to have the option to use pulleys and smaller cable drums to increase the weight the system can hold and also I’d worry about a thinner cable cutting into the drums.

As the cable winds onto the drum it will stack on top of itself and a change of even 2mm, 100mm to 102mm, diameter on the drum will be about a 100g change in the counterbalance. The carriage supported is on rails using bearings so the friction is low and I worry that this will be enough to put the whole thing out of balance.

I spent some time researching counterbalance design using these spring motors but there's not much out there so I thought I come here and see first of all if I’ve got this right and second if there is some solution someone can see that I’m just not thinking of.

This is the spring: https://www.ondrives.com/sr116

This is an example of an assembly: https://www.ondrives.com/ba274-s

Any advice appreciated.

 

[JBA]

Are you asking about suggestions for a different type of counterbalance system; or a modification to eliminate the force differential problem with the current constant torque motor.
Note: Based upon the 25 revolution limit of your motor, if you use a multiple pulley arrangement you will need to increase the cable drum size in order to maintain your current travel distance and that will reduce the pulling load of the motor so you will end up with no force gain for your effort.

 

[pete]

Yes, I'd have to lose travel to gain force but that's something that I'd like to be able to do.

I'm looking for a solution to the problem of the varying force due to the cables building up on the drum or another counterbalance option I guess but I'm not sure what else would have the travel or adaptability.

 

[Tom.G]

How about just adding some friction to mask the torque variation and use a slightly stronger spring or different size drum?

 

[pete]

Adding friction would work but I really want to keep the friction low, it's important it moves freely.

It's relative to the drum size, as the diameter increases the force for any variation on the diameter decreases. Stronger springs help but these are expensive and heavy, you'd improve the situation by tripling the weight and cost but not solve it.

Using pulleys is the best I can think of, this will increase the load/diameter at the sacrifice of travel and I can lose travel.

It just feels like bad design. The problem is never fixed just improved. I've been trying to come up with a way to implement some kind of storage drum feeding cable to the drum on the output shaft but have come up with nothing that could work.

 

[JBA]

What is the source of the load you are trying to counterbalance i.e. is it the weight of a vertical load, a load traveling on a sloped surface or a mechanical load of some type?

 

[Tom.G]

Two other possibilities come to mind, but they increase size and/or complexity.
1) Make the take-up drum long enough to wind the cable in a single layer
2) Use two drums and motors. The drum closest to the load will do most of the work and have a few turns of cable around it, and the 'free' (retracted) end of the cable will be taken up by a second drum.

In approach 2),
If the cable slips on the first drum, either add more turns around it or increase the tension from the second drum.
If side-to-side wandering of cable on the first drum is a problem, try giving it a slight hourglass shape... or adding an ordinary pulley downstream of it.

Cheers,
Tom

 

[pete]

I will have to make something up and test it to see exactly what the problem looks like, maybe it will lay down in a neat row every time naturally, but I doubt it.

The diameter of the take-up drum dictates the counterbalance load so is fixed.

If the cable moves from one side to the other without passing over itself then at a width of just over 20mm its possible but how to feed the cable onto the drum in such a way it always does this?

I spent some time looking at drum profiles and idlers and such but all the ideas I got down either looked unconvincing or became too complex. I can find no standard solutions for this, only in much larger cable drums where a dedicated mechanized guide idler is used.

My first solution is to put an idler between the load and the drum. I'll put some thread on its shaft with a pitch of 1mm then connect it to the output shaft from the drum so they both turn 1:1 and the threaded connection should then move the idler 1mm per turn along its shaft. I need to draw this up but it looks promising.

The second option is a second drum. A storage drum as you say with its own small spring to keep the cable under tension passing a few loops around the output drum to set the load. But Dyneema is slippy stuff and I'd worry that by the time you'd made enough loops to make it secure you'd be back to the same problem, though admittedly in a much-reduced way. As the output drums are changeable to change the counterbalance load it would also make changing the drums very fiddley.

However, If I replaced the cable with a timing belt. I'll have to look it up but I'm pretty sure something like a HTD 5M 9mm wide belt can happily carry the 20kg max load, then your idea of the second storage drum could feed changeable pulleys. Never slips, no backlash and off the shelf parts. I could get the needed 8 meters onto something like a 170mm diameter drum.

I think both could be workable solutions.

 

[JBA]

While it is more complex, one solution to control the wrapping for long single wrap drum would be to use the type of line guide that is used to control line wrapping on fishing reels. You have a fixed reciprocating system with a single wrap on the drum, so you could simplify that mechanism by using a screw thread on a free turning shaft that receives and feeds the line to your power drum. It could have one initiating wrap and tension from your load on the line would keep it in the groove.

 

[Tom.G]

To get a cable to take-up in a single layer, the key is the approach angle of the cable to the drum. For instance if the cable has a fixed guide/constraint close to the drum, it will readily wrap over itself. Move the guide further from the drum and it tends to wrap in a single layer.

I've never bothered any detailed calculations but I expect controlling variables to be:

• Approach angle
• Cable diameter
• Cable stiffness
• Cable coefficient of friction (to itself)
• Drum diameter

Cheers,
Tom

 

[pete]

Sorry about the delay. I had to go away there briefly. I had a close look at the spinning reels. amazing design. A guide on a threaded shaft with the pitch matched to the diameter of the thread is the idea I'm working on at present.
The variables for the test are appreciated too, I hadn't thought about the distance being an important factor.

 

[Baluncore]

There is a similar device. A “tool balancer” employs a fusee pulley to compensate for the spring constant.
You can hang one balancer from another to get a greater working height.
google; tool balancer

 

[Baluncore]

A “hydraulic jigger” employs two multi-sheeve pulley blocks to store a long length of rope. The pulleys are pushed apart by a hydraulic ram. The same technique could be used for your counterbalance system. As an example, 5 pulleys per block gives you 10 falls, with a 10:1 ratio. 5 Kgf end tension needs 50 kgf between the pulleys. To get 10 m movement you only need 1 m change between the pulleys.
https://en.wikipedia.org/wiki/Hydraulic_jigger
“Roller blind chain” has spherical beads equally spaced along a flexible cord. (Cheap on eBay). It runs over a pulley that has peripheral holes along a central slot for the cord . Use a low speed DC electric motor to torque that pulley. Run a constant current through the motor to maintain constant torque. Adjust the current to adjust the tension. Use a similar motor driven take-up spool, or a chain locker to store the unused end of the chain. The take-up spool tension should be minimised so it does not significantly reduce the controlled tension.

 

[pete]

A Fuse pully accommodates for fall off in spring tension as I understand but it did give me an idea. As the cable is so thin I could cut a helical groove from one end of the barrel to the other for the cable to sit in. Do it the same way you'd thread on a lathe but with a tiny "U" groove form tool. Or maybe not, maybe a standard threading tool, the V groove may hold the cable better under tension. Also the two-barrel idea, storage and output with the groove on the output for just 3 or 4 turns.

 

[pete]

I'll have a look into the hydraulic jigger but I must say it sounds complicated. I don't want to have motors or electronics in this design. I'm curious about the pully system thought.

I did think of the bead chain and pully for another design some time ago. I found metal ones but when I looked into it I found that they are all basically decorative and don't take a load of any kind.

 

[Baluncore]

I found metal ones but when I looked into it I found that they are all basically decorative and don't take a load of any kind.

They take the load of raising the blind, by human hand.
Traditional "ball chain" is metal and can be stronger than plastic ball on cord.

The hydraulic jigger can have a fixed upper block with a heavy lower block hanging below. There is no need for a spring or hydraulics.

 

[pete]

They do but I have a 20kg load at the top end. There are larger ones but I doubt ball chain would outperform Dyneema or timing belts. It seems to be marketed as a lanyard or retention chain of some kind. My other concern is noise, I want the mechanism to be silent.

That aside the mechanism you are describing inside this jigger sounds interesting, Wiki describes it as pulleys in reverse but I don't really understand where the force is coming from and can find very little on it online to help me understand the mechanism.

I have to get to work. I'll have a proper look when I get home.

 

[Baluncore]

The diagrams on the left of the page demonstrate the "jigger" concept.
https://en.wikipedia.org/wiki/Block_and_tackle#Mechanical_advantage

 

[pete]

I can find no decent description of how this jigger works online but I think it's a set of pulley blocks with a hydraulic ram between them replacing the weight.
This is an interesting idea but to bring the device down to a manageable size my travell couldn't be beyond about 30cm. It still might be a workable idea though, I'll have to draw it up.
My second concern would be how constant is the force out of a hydraulic spring?

I'm also a little lost as to how using two fuse pulleys dose anything. If a fuse pully evens out the force from a spring by a gradient on the diameter how would two of them help apart from just putting the fall off in the force back into the output?

 

[Baluncore]

I can find no decent description of how this jigger works online but I think it's a set of pulley blocks with a hydraulic ram between them replacing the weight.

That is correct. A hydraulic jigger uses hydraulic pressure to force the blocks apart.

My second concern would be how constant is the force out of a hydraulic spring?

That will depend on the linearity of the spring. If you use an air-spring or gas-strut it will have a reciprocal force with length. If instead you pull a vacuum in a cylinder, it will present a linear force with length.

To use only 30 cm block input, to get 10 m cord output, you would need a cord that passes 10 / 0.3 = 34 times between the blocks. That will require 17 pulleys on each block. For 5 kgf output the force between the blocks would need to be 34 * 5 kg = 170 kgf.

I'm also a little lost as to how using two fuse pulleys dose anything.

The term is fusee with double ee. The limitation on the fusee pulley is the length of the spiral path that stores the retracted cord. When a tool balancer is needed with a longer fall than is available, two balancers can be used in series, one hung from the other.

 

[pete]

I understand the fusee pully now.

I'm confused, I know nothing about hydraulics but given that there must be a sealed cylinder as it needs to maintain pressure then I would think compressing or expanding air would give a gradient in the force along the length of the spring. How could it give a constant force if the pressure changes with the size of the cilinder? Is there a name of a type of hydraulic spring that gives a constant force so I could google it?

Is there any type of hydraulic spring that could be adjusted, there's an engine hoist in the workshop I use that uses a lever and release valve the jack-up and release the pressure in the cylinder. A spring that could be adjusted like this would be amazing for weight and adaptability but again I can't see how it would be constant in the force along the stroke.

 

[Baluncore]

I'm confused, I know nothing about hydraulics but given that there must be a sealed cylinder as it needs to maintain pressure then I would think compressing or expanding air would give a gradient in the force along the length of the spring. How could it give a constant force if the pressure changes with the size of the cilinder?

That will depend on the linearity of the spring. If you use an air-spring or gas-strut it will have a reciprocal force with length. If instead you pull a vacuum in a cylinder, it will present a linear force with length.

Is there a name of a type of hydraulic spring that gives a constant force so I could google it?

Yes, it is called a vacuum. In effect it is the lack of atmospheric pressure.
When you pull a vacuum in a sealed cylinder, the force is constant with position.

 

[pete]

I hadn't really thought of it like that. I guess I just thought by vacuum you meant low pressure, so increasing the chamber would lower it still and decreasing the chamber would increase it but if there's nothing there at all then that means the pull would be the same no matter how much you increase it, is that right?

That's cooked my noodle a bit. Does that mean that if you had a completely sealed cylinder of infinite length you could just keep sliding a piston increasing the volume of the canister and it would remain at the same constant force no matter how large it got? Seems odd to think of increasing a space without putting something in there.

I suppose in the vacuum of space you could endlessly increase the space inside without applying any force at all, how odd.

If you can think of another name for it or a link. I had more luck Googling the Jigger design .

 

[Baluncore]

I suppose in the vacuum of space you could endlessly increase the space inside without applying any force at all, how odd.

Another advantage of doing it in space is that cylinder wall differential pressure is zero, so the cylinder walls, and piston seals, can be virtual and imaginary.

Does that mean that if you had a completely sealed cylinder of infinite length you could just keep sliding a piston increasing the volume of the canister and it would remain at the same constant force no matter how large it got?

That is correct within limits. Such a technique is limited by atmospheric pressure. It also assumes that the atmosphere volume is >>> the cylinder volume.
You can approximate it by having a large gas reservoir, (accumulator), connected to a smaller hydraulic cylinder, with hydraulic oil filling the cylinder and connection line only.

Also consider the mass of the cord. With a fusee the weight of hanging cord increases as it is withdrawn from the fusee. That partly compensates for the spring constant. The jigger cannot do that if it employs a weight to separate the blocks, but it can if the cord is vertical and the separation is forced by a cylinder and accumulator.

If you can think of another name for it or a link. I had more luck Googling the Jigger design .

For vacuum springs you need to go back to first principles, they do not appear in texts that I know of. I have seen them being considered for energy storage in constant pressure hydraulic systems, 40 years ago.
The best information on jiggers comes from 100 year old engineering texts.

 

[pete]

A brief contemplation of designing a fusee pully for any given system, especially one like this where the output can be adopted by the user for different loads and lengths.
Just the thought of trying to work out the circumference... I don't even know the word for it, the length of the path on a tapered spiral pully. Never mind the torque gradient.
Now I've written it down it does seem vaguely possible but far beyond my half recalled GCSE maths. This is without mentioning fabrication. Though I do half recall a way to thread tappers on a lathe.

I love a challenge but remember I'm on here because I can't work out how to get the cable on a drum wrap over its self. This project, like every project, has already spiraled out of control. If my better half finds what you're written she'll have words.

Don't get me wrong, it's fascinating stuff and I'm glad you introduced me to the jigger and fusee pully and the weird idea of expanding vacuum chambers but as a practical solution the Jigger or method of putting a spring between a set of pulley blocks needs a mechanical spring with a constant output even if it's only over 30cm and inventing something I bearly understand will be a stretch.