How can I safely vary the current to a variable torque DC motor for my project?

[MacLaddy]

I have recently learned that in order to vary torque output from a DC motor, I need to vary the input current.
A project I am working on requires that I have a DC motor that will be backdriven while providing resistance to a lever. A max torque of about 10 lbft will be provided from the lever. (that's about a 1.5 factor of safety)
Do any of you know of a safe way of varying current to a DC motor for this purpose? Can it be accomplished with an arduino? Would this just fry the DC motor being backdriven?

Any information is appreciated.
Mac

 

[NascentOxygen]

I have a DC motor that will be backdriven

What do you mean by "backdriven"?

 

[MacLaddy]

What do you mean by "backdriven"?

To drive the motor in the opposite direction that it is trying to go with the intention of causing resistance to the lever. Very similar concept to what is seen in a force feedback joystick.

 

[RNickl]

Sounds like a dynamic brake. Sure you don't want a resistor?

 

[NascentOxygen]

In principle, I don't think there is any prohibition, but you must ensure the current is limited to a safe level. What's a safe level? If it's a DC motor with brushes, and the rotor isn't turning, then there will be problems with one set of commutator segments and winding getting all the power and having no time to cool down. Because it isn't turning, the fan won't be moving air over the rotor so heat buildup will be a major consideration: I think you'll need to devalue the motor's rating massively.

 

[MacLaddy]

Sounds like a dynamic brake. Sure you don't want a resistor?

I'm not familiar with a dynamic brake. Can you give me some more information about that?

 

[MacLaddy]

In principle, I don't think there is any prohibition, but you must ensure the current is limited to a safe level. What's a safe level? If it's a DC motor with brushes, and the rotor isn't turning, then there will be problems with one set of commutator segments and winding getting all the power and having no time to cool down. Because it isn't turning, the fan won't be moving air over the rotor so heat buildup will be a major consideration: I think you'll need to devalue the motor's rating massively.

Good point, I hadn't thought about the brushes. Do they make small variable speed DC motors that are brushless?

 

[Nidum]

Have you considered stepper motors ?

They can be set up to give many different response characteristics and they will work under load at dead slow and zero speed for long periods without damage if you keep the phase currents within the specified limits for the chosen motor .

 

[Windadct]

To me "a lever" implies a relatively small range of motion - are you saying the motor will be at near zero speed? If so the max current could be limited by just the Winding resistance. Also - the number of poles of the motor will affect how "smooth" the torque can be applied, a Brushless DC (BLDC) motor can do this, often called a servo-motor, and they are used in robotics for just this high torque, zero speed case. But still they are typically connected through a gearbox.

Maybe I am over-reading this but this seems like it may be very simple or very complicated.

 

[NTL2009]

If you are just trying to provide a force against a lever, a rotary motor is complicating things, and not optimal for the reasons given - a motor wants to spin (unless you use a stepper motor, but those are more complicated to control). Why not use a simple solenoid (likely cheaper, and easier to implement)? You can think of a solenoid as a simple, linear motor. It is designed for static operation like this, and is easy to control with current (withing its rated limit).

 

[MacLaddy]

Have you considered stepper motors ?

They can be set up to give many different response characteristics and they will work under load at dead slow and zero speed for long periods without damage if you keep the phase currents within the specified limits for the chosen motor .

We thought about using steppers, but don't want to deal with the control issues that are involved.

To me "a lever" implies a relatively small range of motion - are you saying the motor will be at near zero speed? If so the max current could be limited by just the Winding resistance. Also - the number of poles of the motor will affect how "smooth" the torque can be applied, a Brushless DC (BLDC) motor can do this, often called a servo-motor, and they are used in robotics for just this high torque, zero speed case. But still they are typically connected through a gearbox.

Maybe I am over-reading this but this seems like it may be very simple or very complicated.

It probably is quite simple, but I'm over complicating it. The lever is simply hand operated, but needs a force profile. Imagine a backhoe lever, or a forklift. It is similar to that. As above, we don't really want to use a servo motor. Our newest thoughts actually align with the below quote.

If you are just trying to provide a force against a lever, a rotary motor is complicating things, and not optimal for the reasons given - a motor wants to spin (unless you use a stepper motor, but those are more complicated to control). Why not use a simple solenoid (likely cheaper, and easier to implement)? You can think of a solenoid as a simple, linear motor. It is designed for static operation like this, and is easy to control with current (withing its rated limit).

Yes, I actually just thought of this earlier today. (wish I had seen your post sooner) I actually just purchased a solenoid that is capable of 25N, but only over a 10mm stroke. This might be okay for a demonstration, but I will need something bigger. I'm guessing the stroke of the lever will be about 45°, perhaps a bit less. At 7" tall and with a max pull of 6lb, it puts about 42lb-in of torque at the shaft. If I can manage to get 3 inches of leverage at the bottom (iffy), then I need at least 14lbf (or ~20N) at the solenoid. However, the stroke length would need to be about 2.4 inches.

Does anybody know where I can get a solenoid that can meet the characteristics described above?

Thanks everyone,
Mac

 

[MacLaddy]

One additional question...and this may be dangerous and/or dumb...feel free to say so if it is. But if I take this solenoid that I just purchased.

https://www.amazon.com/gp/product/B016MMP374/?tag=pfamazon01-20

And I replace the core of it with a longer piece, then I will achieve a longer stroke...right? I'm not an electrical guy so please forgive my ignorance. But if it's longer then it will just keep inducing current until the object is past the center of the solenoid. Is that right?

 

[Tom.G]

replace the core of it with a longer piece, then I will achieve a longer stroke...right?

Right!
But the force versus displacement will be different. If it works for ya', that's probably the cheapest solution.

 

[MacLaddy]

Right!
But the force versus displacement will be different. If it works for ya', that's probably the cheapest solution.

If the force on a plunger inside of a solenoid is maximum at the maximum distance, then would making the plunger longer create a larger force?
From what I'm seeing, the force from a solenoid is dependent only on the length of the solenoid...not the plunger. I'm a bit out of practice with the physics for this, so I could be very wrong.
And yes, I just typed two things that completely contradict each other. Chances are better that I'll at least get one right that way.

Mac

 

[Tom.G]

...the force versus displacement will be different.

If the force on a plunger inside of a solenoid is maximum at the maximum distance, then would making the plunger longer create a larger force?
From what I'm seeing, the force from a solenoid is dependent only on the length of the solenoid...not the plunger. I'm a bit out of practice with the physics for this, so I could be very wrong.

That's two of us that could be wrong.

"...The force from a solenoid is dependent only on the length of the solenoid...". Lacking specific knowledge, I will accept that as fact.
With a longer plunger, then the initial forces will be identical. The force gradient then has a shallower slope to the center (zero force) position.

(I also agree that my post #13 could have been clearer. (another case of 'after midnite thinking'!))

 

[Merlin3189]

Now that we're getting down to a solenoid or stalled motor, I wonder whether you could simply use permanent magnets! You don't need to provide any power to maintain the force and they don't get hot.
The question is, what sort of force vs position profile you want. Presumably you don't simply want a force proportional to displacement, or you'd be using a spring!
The only thing I can deduce from your posts so far is that you seem to want constant torque, though you did mention needing a force profile. Perhaps it would help if you clarified what that profile is exactly.

 

[MacLaddy]

Now that we're getting down to a solenoid or stalled motor, I wonder whether you could simply use permanent magnets! You don't need to provide any power to maintain the force and they don't get hot.
The question is, what sort of force vs position profile you want. Presumably you don't simply want a force proportional to displacement, or you'd be using a spring!
The only thing I can deduce from your posts so far is that you seem to want constant torque, though you did mention needing a force profile. Perhaps it would help if you clarified what that profile is exactly.

I can't share the force profile (literally, I signed an NDA), but I can say that it looks almost like a 4th degree polynomial along a stroke length vs. feedback force graph. It needs to be programmable for other profiles as well. Just imagine x-axis as the stroke length, and y-axis as the force. And a funky curve going across it.
The torque I have mentioned is max torque. The profile changes between about 2lb and 6lb of force, or approximately 14 and 42 lb-in.

Because of that profile I don't think a permanent magnet would work, unless there is a way to use it that I am not aware of?

 

[MacLaddy]

That's two of us that could be wrong.

"...The force from a solenoid is dependent only on the length of the solenoid...". Lacking specific knowledge, I will accept that as fact.
With a longer plunger, then the initial forces will be identical. The force gradient then has a shallower slope to the center (zero force) position.

(I also agree that my post #13 could have been clearer. (another case of 'after midnite thinking'!))

Ahh, I see what you're saying. I often see springs on solenoids that are non-linear in shape. Is that to offset this gradient and make it uniform across the stroke length?

**EDIT** That doesn't make sense, actually, unless they aren't compression springs--which, I believe, they are.

 

[CraigHB]

Is there a reason you need to use variable current? It's a lot easier in terms of electronics to use PWM. That being the case you can simply program a solenoid force profile in terms of duty cycle with feedback from lever position.

 

[MacLaddy]

Is there a reason you need to use variable current? It's a lot easier in terms of electronics to use PWM. That being the case you can simply program a solenoid force profile in terms of duty cycle with feedback from lever position.

No, not at all. That's just where I was when I started this thread. If we used PWM wouldn't it need to be an A/C system? I'm a mechanical guy, so I'm only vaguely familiar with the concepts.

 

[CraigHB]

It comes down to power, you can vary voltage, current, or duty cycle with PWM to vary power output. Electronically PWM is easiest to implement. It happens to work especially well with inductive loads like motors, transformers, and of course solenoids.

 

[MacLaddy]

It comes down to power, you can vary voltage, current, or duty cycle with PWM to vary power output. Electronically PWM is easiest to implement. It happens to work especially well with inductive loads like motors, transformers, and of course solenoids.

Can it be implemented via arduino? We had thought about PWM, but only in terms of servo motors--which we rejected.

 

[CraigHB]

Of course, the ATMega MCUs typically used with Arduino (and most MCUs for that matter) have a PWM module just for that purpose.

 

[MacLaddy]

Of course, the ATMega MCUs typically used with Arduino (and most MCUs for that matter) have a PWM module just for that purpose.

Nice, I think we will definitely look into that. I've ordered a solenoid, and we are picking a small (read cheap) DC motor to use. Once we settle on a system I'll probably need some additional advice on this.
I appreciate all the help.

 

[jim hardy]

You've not mentioned a size constraint.

Constant current is easy to make with analog circuits provided you don't need a lot of it.

Is this a professional design or a hobby experiment ?

A plain vanilla shunt wound DC motor makes torque in proportion to product of armature current and field strength.
Permanent magnet ones have constant field strength. Automobiles are full of them, windshield wipers windows seats .
Series wound DC motors have same current in field and armature so torque will be in proportion to square of current.

That'd lend curvature to your torque versus amps function.

approximately 14 and 42 lb-in.

roughly 1 to 3½ foot-pounds?

Have you freedom to experiment a bit? And a junkyard nearby ?
Many lawnmowers have small high torque DC starter motors, some in my junkpile are permanent magnet type.
Homeowner type electric drills and circular saws have series wound "universal" motors that'll run on either AC or DC. If your applied current is small enough that the motor doesn't get hot it'll be safe. Since it'll be stalled it won't get any cooling air , so figure out some way to check winding temperature.

Do any of you know of a safe way of varying current to a DC motor for this purpose?

a search on DC Motor Drive Circuit images produces lots of schematics.

 

[MacLaddy]

You've not mentioned a size constraint.

Constant current is easy to make with analog circuits provided you don't need a lot of it.

Is this a professional design or a hobby experiment ?

A plain vanilla shunt wound DC motor makes torque in proportion to product of armature current and field strength.
Permanent magnet ones have constant field strength. Automobiles are full of them, windshield wipers windows seats .
Series wound DC motors have same current in field and armature so torque will be in proportion to square of current.

That'd lend curvature to your torque versus amps function.

roughly 1 to 3½ foot-pounds?

Have you freedom to experiment a bit? And a junkyard nearby ?
Many lawnmowers have small high torque DC starter motors, some in my junkpile are permanent magnet type.
Homeowner type electric drills and circular saws have series wound "universal" motors that'll run on either AC or DC. If your applied current is small enough that the motor doesn't get hot it'll be safe. Since it'll be stalled it won't get any cooling air , so figure out some way to check winding temperature.
a search on DC Motor Drive Circuit images produces lots of schematics.

Hello Jim,

I apologize for the response time. I didn't realize there was a new reply here.

This is a professional design in the sense that it is a capstone project. I am being sponsored by a local company, but it is academic in nature.
Regarding size, we really can't go much larger then a Nema17 type stepper motor. That is, in fact, where we are right now. We concluded (although with little information) that these motors would have the highest holding torque, and the feedback could be programmed via PWM.

Our torque numbers have also changed considerably. I don't remember the calculations off hand, but they are quite a bit higher then originally expected. However, there is linkage available to turn the rotational motion of the handle to linear motion on a rack, and then gear from there.

I still welcome any thoughts. We are moving forward, but it is a bit of a shot in the dark. This "group" was originally supposed to have two EE's included, but the ME department is so much larger that it just ended up with a bunch of ME's trying to learn as quickly as possible.

Thanks,
Mac