[originator]

I am currently using a Japan Servo and Dy adic controller. The Nema 23 has a 10:1 gearbox on it. The application is not CNC, but a single motor in a linear motion application, moving a 100lb load horizontally on a rail and slide via timing belt and pulley. The speeds are slow, and require an ultra smooth ramp in and out of the curve. The servos provides no vibration, smooth as silk motion.

The system I am using including Nema 23 servo, gear box and controller is running around $800 per system, which is cheaper by far than anything comparable as far as servos. Automation direct is way higher for less torque, and far less options on the controller.

Is there a way to get ultra smooth motion out of a stepper(not using gearing--the cost is too high), plus very low noise? The servo I am using is dead quite, but the steppers I have on my CNC mill(closed loop Lin enginerring nema 23's) are very loud, even at stand still they ring so loud people can't stand to be around when they are sitting idle.

The bottom line is I am looking to cut costs, but don't want to sacrifice power, noise levels, ease of use with a controller with 16 programmable positions, each with independently programmed speeds, accel/decel, torque.

Does something exist in the stepper world that compares to what I am describing, but at a lost less cost?

I have been talking to Lin design, they have some options, but cant guarantee the power I need in Nema 23 without using a gearbox. By the time they add the gearbox, motor and controller I am back at the servo price range.

Thanks for any ideas.

[NC Cams]

Welcome to what I call "the perverse nature of inannimate objects" axiom of life....

A stepper and a servo motor are ESSENTIALLY the same thing, They both consecutively activate a magnetic field in order to turn electrical energy into mechanical energy. The "consecutive switching" of the electromag fields is done via a process called communtation.

As the fields are actuated, they simultaneously push away from and attract to the magnetic fields in the stator. The amount of fields and the overlapping thereof in the rotor will determine the smoothness of the rotor as it commutates.

Using a simple issustration, model trains, they struggled for years to get smooth low speed motion with 3 pole armatures (horrible "cogging" a 6 distinct pulses per rev). Subsequently, 5 pole arms came into vogue - more pulses BUT they overlapped more hence the cogging was less perceptible. The motor makers the learned to "skew" the armature stacks. Rather than being aligned ( I I ) the were skewed ( / / ).

Since the mag field was skewed as well, the cogging was reduced more to the point is is imperceptable. 5 pole motor plus worm gear drives and ultra slow motion with no noticeable cogging/jerking of the trains.

Knowing the above, it breaks down to the issue of how many poles you can switch off and on and can it be done at the speed you want the motor to run or how much gear do you throw at it to cover up any cogging you may have??? Life is full of compromises.

This is where brushed motors do neat things and do so so simply, crudely yet elegantly. They provide dynamically stable, high speed switching capability of high power devices with relatively cheap componentry - brushes. To do the same commutation at comparable speeds AND SMOOTHNESS, requires quite complex and FAST circuitry that will switch comparable power.

Switching power with brushes is pretty simple. Switching high power with transistors has gotten less costly but still can get comparatively more expensive with transistors. IF we can get past this hurdle, we then get to the motor/power dilemma.

If you can't make sufficient power with the motor, you have to make it up with gearing. If you don't have the torque in the motor, you have to make it up with gearing. If you don't have the net speed you need, you need to make it up with gearing but that requires a more powerful motor to drive any speed multiplying gearing.

The quandry becomes self evident. You need either the motor to supply power at the speed you want/need or you need a gear box to multiply it.

At some point, a certain device will provide the acceptabe cost/benefit/performance ratio. At another point, it fails to do so and you have to move on to the next higher element on the food chain capable of doing what you want to do at the speed you need with the required power.

Unless you can find a smooth operating high speed stepper that acts with the smoothness of a self commutating motor (ala floppy disk drive) that has the necessary power (do they make a floppy drive that big?), you're faced with the same dilemma that has face and remain unsolved ever since CNC systems were concieved - namely, how do you come up with a servo system that will perform at the cost/simplicity of a stepper???

As the seemingly immortal Mick Jagger once said, "..you can't always get what you waaaant....".