Preventing Motor Stalling

[Charlie675]

Our team has an arm driven by two cim motors. The motors are obviously strong enough because the arm can fly up in a fraction of a second (no one dead yet!). but we need to hold the arm in a constant position to score. when the motors are set to a value that holds them in a specific place, they work for a few seconds and then the arm suddenly drops. I assume this is because the motors aren't meant to apply torque without moving but I don't know. what should we do? would pulsing the motors help? if so why and what is the best curve to give the speed controllers?
Thanks for the help. our whole design is based on this working!

[ttakashima]

Your arm shouldn't back drive with 2 CIMs in place! If it is you should consider gearing it down a lot more! How long does it take your arm to rotate 90 Degrees?

Or use the Jaguar/Victor in Brake mode where it shorts the motor and prevents it from back driving (I do not know how this is done in code however).

You could also use some surgical tubing or springs to balance your arm and make it "Weightless" or as near to that as you can get.

[114Klaatu0x72]

Hellos. On our robot, we have a winch powered by a fisher price. Unlike cims, stalling them at all generally destroys them rather quickly. We decided on using a bicycle disk brake, driven by a single pneumatic cylinder, to lock the winch into position. Because getting the arm into position is not your problem, but keeping it there is, some type of locking mechanism would seem to be ideal.

Moving on to some math, lets say that a bike may weigh 15 kg, and an average rider may weigh 65 kg, with wheels of a .4 m radius, disc brakes are designed to take about 32 Nm of torque, or about 65 lbs on a 3 foot lever arm. While this leaves out various tire dynamics, among other things, unless you have some crazy manipulator, one of those should more than suit your need.
As long as you have access to a drill press, or even a corded drill in steady hands, and some extra shaft, it should be pretty easy to make a mount for the rotor (the disc part of the break). As for securing the mount onto the shaft, you should either used keyed shaft with a keyed mount, or drilling through both the shaft and mount along their diameters and attaching them with either a roll or cotter pin, as unless you have access to some decent machining tools, attaching the mount in a secure manner will be quite a pain.

Good luck and I hope this helps.

[Mike Betts]

Originally Posted by Charlie675

Our team has an arm driven by two cim motors. The motors are obviously strong enough because the arm can fly up in a fraction of a second (no one dead yet!). but we need to hold the arm in a constant position to score. when the motors are set to a value that holds them in a specific place, they work for a few seconds and then the arm suddenly drops. I assume this is because the motors aren't meant to apply torque without moving but I don't know. what should we do? would pulsing the motors help? if so why and what is the best curve to give the speed controllers?
Thanks for the help. our whole design is based on this working!

Charlie,

Let's look at this from a control-ability standpoint. In order to achieve precise arm positioning for the human driver, you want a rotational speed of about 45 degrees per second. This gives you a gear ratio of about 350:1 at the peak power point.

At that gear ratio, you get over 600 ft*lbf of torque at the shoulder and at the peak power point. Neglecting the weight of the arm, you could lift about 125 lbs at a distance of 5 feet!

In my opinion, you are trying to hang a picture with a sledge hammer.

Regards,

Mike

[mathking]

We are also using a bicycle disc brake for our arm, but an alternative breaking system could be used. In 2005 we had a simple but effective breaking system. A very short stroke pneumatic attached to a heavy piece of rubber partially wrapped the output shaft from the gear box (actually wrapped around a cam-like device on the shaft). When the pneumatic was extended the system could move freely. When the pneumatic retracted the arm was locked in place.

I will also second Mike's point that you probably have way more power than you need to move the arm.

[Ether]

Originally Posted by Charlie675

Our team has an arm driven by two cim motors. The motors are obviously strong enough because the arm can fly up in a fraction of a second (no one dead yet!).

This high speed indicates the motors are not sufficiently geared down.

but we need to hold the arm in a constant position to score. when the motors are set to a value that holds them in a specific place, they work for a few seconds and then the arm suddenly drops.

If you are using Jaguars, they are probably shutting down to prevent damage.

I assume this is because the motors aren't meant to apply torque without moving

Yes, and the Jags aren't meant to supply that level of sustained current.

what should we do?

Gear the motors properly. ideally, you want to select a gear ratio that provides about 45 to 90 degrees per second of arm rotation at a motor speed of half its free speed.

would pulsing the motors help?

No.

our whole design is based on this working!

You have at least 3 options:

1) add a brake to hold the arm in position so you don't have to apply high levels of sustained current

2) add a spring or surgical tubing to support some of the weight of the arm

3) gear the motor properly

Originally Posted by ttakashima

Or use the Jaguar/Victor in Brake mode where it shorts the motor and prevents it from back driving

The "brake" mode you refer to is dynamic braking, not static braking. Without improving the gear ration this is not likely to help.

[Charlie675]

Thanks for the advice, that is unfortunate that our design is flawed. I guess it's better finding this out 6 days before ship date than 2 days before ship date! We'll dry gearing it down more and using a torsion spring.