stalling a DC motor: what will happen?

[Hank Fletcher]

For my application of building a recorder-playing robot I'm currently using solenoids to simulate the fingers, but after having a crack at it I'm beginning to think that using some small electric motors might be a better way to go.

My idea is to use something like this:
https://www.goldmine-elec-products.com/prodinfo.asp?number=G9331
for each finger by attaching a pulley and string to the motor shaft. The motor shaft will rotate almost 360 degrees before it stalls as the finger presses against its closed position on the recorder. The motor will only have to be powered to rotate one way, because at the other end of the string a spring will return the finger (and the motor) to the starting/open position.

My question is: will stalling the motors damage them? My impression is that this is how the steering mechanisms work in low-budget R/C cars, so I thought it might work. Given the stall current and, let's say, about 6V for the motor, is it fair to say that the motor would be able to safely dissipate the power as heat while it's stalled?

 

[Nigel Goodwin]

Given the stall current and, let's say, about 6V for the motor, is it fair to say that the motor would be able to safely dissipate the power as heat while it's stalled?

As long as you seriously limit the stall current, and it's not stalled for very long - stall current might easily be ten times (or more) the full load current. The motor also relies on forced air cooling from it's movement, so if it's not moving it will soon burn out.

 

[dknguyen]

Stalled motors get damaged. The motors that appear to "hold position" are actually being pulsed power and in between pulses they are allowed to droop and then being pulsed again. They are being run far under their stall capability.

 

[HarveyH42]

Pretty sure these are the same motors, I have 10 or so...

https://www.goldmine-elec-products.com/prodinfo.asp?number=G13327

These have nice leads attached, and each has a header socket pin. Don't remember if they had the housing, I used them for jumpers on my AVR programmer. Good torque at 12 volts, takes some effort to stop the bare shaft with your fingers. Little noisey at 12 volts, totally silent at 6, but the torque isn't great, but usable. Two for 99 cents is a good buy. Maybe PWM the 12 volts for slower speed and keep the torque...

 

[Hero999]

You could derate the motor accordingly; for example if you run a 24V motor from only 6V and I doubt it'll be damaged when stalled.

 

[kchriste]

A simple method would be to time how long it takes your motors to make the valves go from open to fully closed. Then add 30-50% to this time and reduce the current to the motor after the timer expires. Should be fairly easy to do if you are using a micro controller to activate the motors and would require no extra hardware to PWM them.

 

[Hank Fletcher]

Thanks, those all look like great tips! I started thinking a bit more about the nature of my application. I need the "fingers" to be able to move fast when they need to, but every now and then any given finger might be pressed down for quite a long time (for instance, on a long note, or when playing a sequence of notes where it happens that a particular finger might have to stay down for all the notes).

So what I'm thinking of doing is making a universal finger that would be a solution to all my present and many of my future requirements in making automated musical instruments. Each finger would consist of two small motors: one that moves the finger into its pressed-down position, and another motor that locks the finger into that position. That way, the pressing finger could hold with as much torque as the motor can provide for a half second or so, and if it needs to remain in that position for longer, the second motor can activate to lock that position. Then the finger could hold down for as long as I wanted, without generating any stall heat/damage or using any power at all.

Sounds a bit complicated, but I think it might be worth giving it a try. Does anyone know of anything like this out there already? I think I've seen some locking solenoids somewhere, but I don't know about anything like I've described.

 

[3v0]

I like kchriste's idea. It uses more electronics but fewer moving parts. It should give you more control over the finger. Full current to get it moving fast and then PWM it to get the sort of action you are looking for. An inexpensive 4 transistor H-bridge which you could build for a buck or two would let you power it both ways.

 

[boxer4]

I don't understand...why won't solenoids work for this application?

Also you could use a current sense, cut off power when current goes up, supposedly when it's reached the end of travel?

 

[Hank Fletcher]

I don't understand...why won't solenoids work for this application?

Also you could use a current sense, cut off power when current goes up, supposedly when it's reached the end of travel?

The solenoids are less efficient - you get less torque/press for power required, and they produce more heat. Also, I'm getting tired of their limitations as far as travel/volts goes.

I can't just cut the power at the end of travel if I'm using a spring mechanism to return to the original position (which I will do, because it's more consistent with the way many keyed wind instruments work, e.g. since the valves on a trumpet already have springs in them, it would be redundant engineering to remove the springs just to be required to replace them with a powered motor or solenoid for the return trip). If I did, the finger would no longer have it's pressing power, but as I wrote earlier, I could lock the finger in it's press position, and presuming it's compressed some sort of sponge against the key, hole, or valve, it will continue to do so until unlocked. The compression aspect is most important in the instance where the finger is covering a hole, since in reality that requires squishing the flesh of your fingers to completely seal the hole.

 

[Krumlink]

The motors that we use in our FIRST robotics usually use about 6amps. When they stall, they skyrocket to 650amps.

BE CAREFUL!

 

[Pommie]

Current can damage a motor in two ways. By creating too much torque causing mechanical failure or by creating too much heat and causing burn out. If you limit the current to 75% (even 100% if the rating is conservative) of the normal running current then you should be able to leave it stalled all day long. This cannot be done with a fixed voltage but requires a constant current source. If you have a good lab power supply you can simulate this easily. Just run the motor unloaded at it's rated voltage and note the current, now set the current limit to 75% of the normal current and stall the motor. It should get warm but will not be damaged.

Mike.

 

[boxer4]

The solenoids are less efficient - you get less torque/press for power required, and they produce more heat. Also, I'm getting tired of their limitations as far as travel/volts goes.

Right. But why is it that pretty much all on/off controls on cars done with solenoids? Designed properly they produce little heat - if you think about it, theoretically once it's in open or closed state, it requires no power to hold it - since it's doing no work! They also can produce a _lot_ of force, the same amount of force as a motor with the same magnet and number of windings. Of course you can gear motors to trade off, but you can do the same with levers for solenoids.

I can't just cut the power at the end of travel if I'm using a spring mechanism to return to the original position (which I will do, because it's more consistent with the way many keyed wind instruments work, e.g. since the valves on a trumpet already have springs in them, it would be redundant engineering to remove the springs just to be required to replace them with a powered motor or solenoid for the return trip). If I did, the finger would no longer have it's pressing power, but as I wrote earlier, I could lock the finger in it's press position, and presuming it's compressed some sort of sponge against the key, hole, or valve, it will continue to do so until unlocked. The compression aspect is most important in the instance where the finger is covering a hole, since in reality that requires squishing the flesh of your fingers to completely seal the hole.

This is yet another method to detect end of travel without using optical methods. But perhaps the easier solution is to just current limit by sticking a resistor in series with the motor, sacrificing a bit of response time and depending on the fact holding a position requires less energy than moving to a position.

I still think a solenoid solution is better unless partial openings are needed (and the response time is a non-issue, free-running motors will be slower). Then again partial openings will require special optical detection anyway unless you use a stepper motor (which still needs rest position detection)...

But if you end up using a stepper motor, it's pretty much a rotary solenoid...

 

[Hero999]

If you limit the current to 75% (even 100% if the rating is conservative) of the normal running current then you should be able to leave it stalled all day long.

True, but I would never limit it to below 100% of the on load rating or it'll have difficulty driving the load, you'll loose both torque and speed.

The best way of protecting a motor from stall is to use PTC resistors (polyswitches) but you might need to cut the power or spin the shaft to get the motor started again and it'll only work on relatively small motors, as I've never seen a polyswitch rated for more than 16A.

 

[Nigel Goodwin]

Perhaps all this explains why it's normally done with pneumatics?

 

[Pommie]

True, but I would never limit it to below 100% of the on load rating or it'll have difficulty driving the load, you'll loose both torque and speed.

If you limit the current to 75% of the normal running current then you will get 75% of the torque and full speed.

I realise now that I made a mistake earlier, I should have stated 75% of the most efficient current. I.E. the current at max power output.

The best way of protecting a motor from stall is to use PTC resistors (polyswitches) but you might need to cut the power or spin the shaft to get the motor started again and it'll only work on relatively small motors, as I've never seen a polyswitch rated for more than 16A.

I don't know what a polyswitch is, perhaps you can describe what it does.

Mike.

 

[Hero999]

https://en.wikipedia.org/wiki/Polyswitch

 

[Pommie]

A polyswitch appears to be a slow acting device, if it's recovery time is quick enough then it could fit the requirements - although, you would need a series resistor to provide a holding current. As the reset time should be less than a second I think it may not be quick enough.

Back to motors, do you agree that at 75% of the most efficient current, you will get full speed and 75% of the most efficient torque? And, you will be able to stall it all day long.

Mike.

 

[Hero999]

A polyswitch appears to be a slow acting device, if it's recovery time is quick enough then it could fit the requirements - although, you would need a series resistor to provide a holding current. As the reset time should be less than a second I think it may not be quick enough.

Why do you need a series resistor?

The only issue is the reset time and whether the current whilst tripped is high enough to move the load, if it's a fan it might be possible that the tripped current is enough to more the get it spinning which would reduce the current enough for it to recover.

Back to motors, do you agree that at 75% of the most efficient current, you will get full speed and 75% of the most efficient torque? And, you will be able to stall it all day long.

Mike.

I don't know, it depends on the motor's torque speed curve besides, limiting the current to 75% of the most efficient current will reduce the efficiency be 75% which wouldn't be a good idea.

I would limit the current to 125% the full load current, perhaps even more, it depends on the torque required to start the load.

 

[Pommie]

Why do you need a series resistor?

You need a series resistor to provide a holding current.

I don't know, it depends on the motor's torque speed curve besides, limiting the current to 75% of the most efficient current will reduce the efficiency be 75% which wouldn't be a good idea.
I would limit the current to 125% the full load current, perhaps even more, it depends on the torque required to start the load.

When a motor is running at max efficiency then the current it draws is the maximum it can draw indefinitely. If we stall the motor then it will not get the cooling effect of the rotating armature and will therefore overheat at that current, hence my suggestion of 75% of the max efficiency current.

Mike.