Closed loop....open loop?

[DAB_Design]

can someone shed some light? Not really understanding the concept, as applied to stepper motors.

[ger21]

MAXNC has some info on there site. They offer th only closed loop stepper system I'm aware of.

Basically, with an open loop stepper system (which most everyone uses), if or when you lose steps, you have no way of knowing it. The machine keeps on cutting away, whether it is in the right position or not.

With a closed loop system, the steppers have encoders, which make sure the stepper is in the right position at all times. If you miss a step (or steps), the software driving the machine will know from the encoder that those steps were not made, and will give the motors more steps to get back into position.

[Splint]

The open loop system simply powers the stepper motor to move each step in the desired direction with no means of "telling" the controller whether the motor has actually moved each step. What can go wrong is when the work load is too much for the motor and it cannot move to the position it is ment to go to, that step is lost and from that point on the program is unknowingly machining incorrectly.
The closed loop system has an encoder mounted on the shaft of the motor which gives a feedback of the position of the shaft, that way an error can be detected and compensated for.

[replicapro]

I thought the closed loop system on the maxnc was only servo driven. I didnt realize you could go closed on steppers. This changes a bunch of factors, because arent steppers more powerfull at low speeds?

[HuFlungDung]

Steppers are more powerful at low speeds than steppers at high speeds, but I doubt that they would have better torque at low speeds than a servo, given equal size rotors with similar current draw characteristics. In most situations, guys will oversize their steppers (relative to closed loop servos applied in the same application) to try to get reliable performance out of an open loop system.

[DAB_Design]

Thanks for the info guys.

So basically, with stepppers, if you're going to be pushing your machine to the limits, closed loop is better. All servos are closed loop? But more expensive.

[metlmunchr]

Servos can be run in open loop via a servo step and direction drive such as the Gecko or Rutex. Several people far more knowledgable than myself (Ballendo for one, IIRC) say attempting to run a closed loop stepper system is not a worthwhile endeavor. If you're losing steps from overloading the axis drives, chances are you'll be making scrap parts regardless of whether you have servos or steppers in open or closed loop. At one time there was a reminder on the AHHA website stating that most machined parts on the first Apollo lunar lander were made on machines driven by open loop steppers. IMO, that says a lot about what can be accomplished with steppers.

[HuFlungDung]

DC Servo motors have no "natural" method of cogging themselves along, they are just DC motors.This is why they require an encoder to measure and monitor their position. If the connections to the encoder are broken, incorrect, or missing, the servo motor just "starts running" when voltage is available, just like any other motor. This is called a "runaway" and looks quite scary when you see your machine running at maximum high speed towards the end of travel

The controller monitors the encoder position. If a new position is commanded, the controller creates an error signal that must be brought to equilibrium in the servo circuit. The servo amplifier actually magnifies this error signal into a voltage pushing enough current that is sufficient to run the motor. As the motor nears position, the encoder feedback to the controller reduces the amount of the position error signal. So the motor ramps down in speed as it approaches the new position, and theoretically "stops on a dime". There is more to it than this because of inertia, and stuff, so there are other tricks being applied by the controller to anticipate motor movements, and to make it accelerate and decelerate at a low enough rate so as not to overload the amplifier, and to stop without overshooting the position and backing up.

[HuFlungDung]

Sorry Metlmunchr, I seem to be following you around today.

My intention was not to try to upstage you. I appreciate your comments and have learned from them myself.

[metlmunchr]

Hu, I've read many of your posts in the past and have learned lots from them. Particluarly since I'm a fellow long term sufferer of bobcad. Like yourself, I'm just here to learn and share knowledge. Glad to make your acquaintance

Cliff

[ESjaavik]

Hu, that's an excellent short-description of a servo.

Now let me try with a simplified visualization of the stepper motor:

It works like an electric ratchet. Think of it like you have a strong spring at the end of the handle, you're pulling at the other end of the spring. One pull at the spring (step input) and it moves one step (ratchet click).

The direction input works like the handle on a ratchet that you flip to go clockwise/anticlockwise.

If the resistance to turn is too great, the spring will not pull the handle, and you will not advance (lost step because too much torque needed).

Then to the closed loop. Say there is an encoder that delivers one pulse per step. It could be used to verify that a commanded step actually takes place. If it doesn't, the needed torque is higher than the step motor can deliver and it will not help sending more step pulses. So the drive electronics cannot bring it back into synchronization. What it could do though is to fail gracefully, notifying the running proogram that it should stop instead of continuing with the wrong assumptions of where the tool is. This would typically be a fault input on the computer running the program. The encoder and electronics doing this would cost money, and so the servo solution will beat it on price.