Stepping motors supply the high torques necessary to directly drive many mechanisms, with most others only requiring a single-stage gear or pulley. Simple systems such as this provide a large safety margin for a long operating life.

But applications requiring a coordinated and gentle motion in a single operation often necessitate using servomotors, which have a low pole-count design and produce relatively low continuous torque at relatively high shaft-speeds. Servomotors thus often require precision gearheads rated to withstand normal and operator-applied loads to provide adequate torque (and reduce the inertial mismatch between motor and load).

How can a designer get the best of both worlds? Using advanced closed-loop algorithms with high pole-count, permanent-magnet AC-servo motors (called stepping motors when operated in open-loop mode) provides smooth motion and a low-cost, simple mechanical design. Doing so also supplies the high-torques necessary to eliminate gearheads in many applications, reducing the cost and size of systems and improving their reliability.

This combination requires a position-feedback sensor and a servo controller that can drive a high pole-count closed-loop motor. Special controllers such as the SilverDust by QuickSilver Controls Inc., San Dimas, Calif., quicksilvercontrols.com, handle a wide range of motors from NEMA 11 up to NEMA 34 frame. The result is a true servo system that eliminates resonance and vibration issues associated with open-loop stepping motors while providing excellent torque and minimal heating.

The controllers use a few principles of motor design. Basically, the torque constant of a motor, defined as the ratio of output torque to input current (Kt=Nm /A), is proportional to the number of magnetic poles, all else being equal. Each electrical cycle of the motor advances the rotor one pair of poles.

So, in comparing a 4-pole (2-pair) to a 100-pole (50-pair) AC servo, rotating the stator field 100 times produces 100/2 or 50 rotations of the shaft in the 4-pole but only 100/50 or 2 rotations in the 100-pole system. When the two motors produce the same mechanical output-power (torque times speed), the 100-pole motor produces 25 times as much torque at 1/25 the speed as the 4-pole motor. The motor thus acts as a 25 to 1 gearhead, an effect commonly referred to as “magnetic gearing“.

Get rid of the gearhead

A recent application that handles a variety of primary-sample tubes and sample cuvettes that range down to pediatric sizes in a medical analyzer illustrates how using such motors in close-loop mode addresses the long life, low maintenance, and decontamination requirements of most clinical instrumentation. The test samples are held in rectangular racks in trays.

The system's loader, which is the interface between the operator and the processor, has two motorized gates positioned in a channel, providing the tray-input mechanism. The operator “requests” to load samples, causing one gate to slide open and the other to move to a holding position. The operator places the input samples between the gates.

When the operator closes the input-queue cover, the system must determine the number of sample racks loaded and then present them one at a time to a transport mechanism that loads them into the instrument. This requires a slight compression force on the racks between the two gates to prevent bouncing. When the tests on a rack are complete, the system feeds the rack back into the tray, near the second gate, which then pushes the completed rack into the output queue, where the operator removes the samples.

The application requires smooth motions to prevent agitation of the spun samples in primary tubes in the absence of a separator. Motions must be consistent over a significant range of loads ranging from a single pediatric sample up to a set of full sample racks loaded with primary sample tubes.

The mechanism must also minimize heating to minimize evaporation, especially important for smaller samples. And it requires controlled forces to prevent the sample tubes breaking and generating aerosols when the mechanism jams. The mechanism must also survive forced movements by an operator without damage.

In this system, the mechanism motion has 2.5 seconds to cover 50 cm, for an average speed of 20 cm/sec. Each gate is configured as a direct belt-drive system using a 2-mm pitch, 6-mm wide, curvilinear neoprene belt with Kevlar reinforcement, and a 20-groove pulley with a 12.73-mm pitch-diameter. The neoprene material was used for its reduced static build-up and natural mechanical-damping characteristics.

Design notes for the belt and pulley combination show better than 0.63 Nm tooth-jump resistance with 4.4 N installed tension, which gives a sufficient margin for A17H-1 motors producing 0.17-Nm torque when operated with the special controllers. The system produces a gate-force greater than that required. Using a reduced motor torque setting allows limiting the available torque to the system requirement. In addition, a thin flywheel adjacent to the pulley raises the effective inertia of the motor and simplifies control-system tuning for smooth motions over a wide range of loads.

To minimize the system's temperature rise, the closed-loop control provided by the special controllers applies only the current needed to make the required motion. In fact, compared to the common 60% safety margin for open-loop stepping motors, which require full winding current while only using a maximum of 40% torque, closed-loop control for the same torque produces only 16% of the full winding-current heating. In most operating systems, the average torque requirement is less, so heating is lower yet.

Also, operated as servos, the motors grip and hold without chatter, whereas open-loop stepping motors chatter when driven against a hard stop, skipping backwards four full steps (or more) before resuming their forward motion. The gripping of sample racks requires the motors to push the gates against a varying load until the load is lightly compressed between them.

Closed-loop operation provides the torque control that lets one gate be dominant (set to a higher torque limit) while the second gate provides the squeezing force (set to a lower torque limit). And the position feedback from each motor continuously reports their positions, which provides a way to calculate the number of sample racks loaded.

The high-inertial mismatch capabilities of the special controllers easily accommodate these widely varying loads. The controllers use both active and passive damping to reduce system resonances, which allows higher system gains for greater bandwidth and tighter motions. The passive damping of the SilverDust, for example, uses a patented drive technique to vary the effective driver impedance as a function of frequency.

With this technique, the driver approximates a current source producing torque-controlled motion at lower frequencies. At higher frequencies, the unit's impedance drops so that it more closely approximates a voltage source. The passive interaction of the motor winding with a voltage drive causes a damping current to be superimposed upon the driving current. The damping current provides a dissipation path for unwanted high-frequency motion, which reduces resonance oscillations, or vibrations.

A second active “synthetic viscous inertial damper“ mechanism uses software in the control loop to approximate the control loop effects of adding a viscous inertial damper to the motor shaft. A real viscous inertial damper typically contains an inertial test mass inside a sealed housing filled with thick oil. At constant speeds, the mass spins at the same speed as the motor shaft, essentially adding only inertia to the system.

But at higher accelerations, the test mass tracks behind the motor shaft, causing a shearing action of the coupling oil. This shearing action dissipates energy, damping the torsional resonances in the system. Such viscous inertial dampers are effective, but often costly and physically large. The software, on the other hand, provides a damper in the control loop, while adding no additional size or cost to the system.

Design notes

The application described is intended to demonstrate design considerations and is actually a composite of several different operational systems.

To see the design notes for the belt and pulley combination, visit http://www.sdp-si.com/D260/PDF/part1.pdf.