Direct Drive Motors

From AltAzScopes

Jump to: navigation, search

Contents

[edit] Requirements

Large diameter, frameless, direct drive motors are available off the shelf. They are, however, somewhat expensive, typically costing several thousand dollars per motor. They are designed to handle large loads at high RPM with excellent efficiency. For those who wish to construct telescopes on a limited budget, these off-the-shelf motors may be overkill for our very low speed, modest torque requirements. We are fortunate, however, that the permanent magnets and coils required to construct a simple, axial flux direct drive motor are quite inexpensive—less than $300 for a motor that is adequate for the 0.7-meter telescope.

The motors must supply sufficient torque at reasonable current to overcome wind gusts and static imbalance while accelerating the telescope axis at the highest desired slew rate. The Nasmyth-focus configuration completely eliminates altitude axis imbalance associated with instruments and eyepieces, easing the direct drive torque requirement for this telescope.

Table 1: Motor Requirements
Maximum Rotational Inertia 16 kg-m^2
Maximum Desired Acceleration 0.25 radians / sec^2
Torque Needed for Acceleration 4 N-m
Torque Needed to Overcome Wind Gusts 8 N-m
Torque Due to Imbalance 2 N-m
Total Torque, including Margin 20 N-m


[edit] Design

The plan view of the motor. The poles of the magnets alternate between North and South. The coils are connected in series in three phases, where the phases are given by the numbers.
The plan view of the motor. The poles of the magnets alternate between North and South. The coils are connected in series in three phases, where the phases are given by the numbers.
A portion of the cross section of the axial-flux motor. The coils are 0.25” thick and the gap is 0.12”.
A portion of the cross section of the axial-flux motor. The coils are 0.25” thick and the gap is 0.12”.

Motor drive electronics are inexpensive for peak currents up to 5 amps, and portable telescope control systems can be powered by two 12-volt batteries in series, so we designed our direct drive motor for 5-amps per phase at 24 volts. Under normal usage conditions of light wind, sidereal tracking, and good balance, the torque needed will be quite low. Thus the DC current draw will also be low, eliminating the need for expensive, high-efficiency motors.

Although radial-flux motors have the highest efficiency and motor constants, they also require stamped and laminated cores and machined rotors and stators. An alternative, lower cost (but lower efficiency) approach is to employ an axial-flux configuration with a large diameter rotor and stator. The large diameter is needed to supply the requisite torque. The 0.7-meter telescope has ample room, so no penalty is incurred for a large-diameter motor. To simplify the motor and its assembly, steel is used only under the magnets; the stator has no magnetic core material.

Table 2: Motor Design and Calculated Performance
Magnets (32) Type N42 NdFeB, 0.5” X 1.0” X 2.0”
Coils (24) 20 AWG 100-turn, L=10 cm / turn
Average Magnetic Field in Coil 0.3 Tesla
Force Constant (per phase) 23 N/amp
Average Radius 0.23 meters
Torque Constant (per phase) 5 N-m/amp
Winding Resistance 5 ohms


[edit] Prototype Fabrication

[edit] Coils

Coil Making.  Three coil forming jigs are shown, as well as three finished coils.
Coil Making. Three coil forming jigs are shown, as well as three finished coils.

Each coil consists of 100 turns of 20 AWG magnet wire. They were made three at at time on the coil forms shown in the figure. Clear silicone adhesive (RTV) was used to hold the shape of the coils before the coils were permanently mounted to the rotor. The RTV was applied after 10, 30, 50, 70, and 90 turns, and a final layer was spread over the outside of the coil after all 100 turns had been wound. The RTV was allowed to cure for about 2 hours before the coil was removed from the form. Wood finishing wax was used as a releasing agent, preventing the RTV from sticking to the form.

[edit] Stator

The completed stator is shown mounted to the test apparatus.  The large bearing is part of the test set-up.
The completed stator is shown mounted to the test apparatus. The large bearing is part of the test set-up.
The wiring channel on the back of the stator.
The wiring channel on the back of the stator.

The stator consists of 24 coils mounted on plywood. The ends of each coil were passed through holes drilled in the plywood in order to keep the wiring tidy. A wiring channel routed into the back of the plywood was used to keep the wiring from interfering with the mounting of the stator. The coils were mounted to the plywood using laminating epoxy. While the epoxy cured, a disk was clamped to the stator to keep the coils as thin as possible and coplanar.

The stator was wired in a three phase configuration. Every third coil was connected in series in a daisy chain fashion. All coils were wired with the same polarity.

[edit] Rotor

The fabricated rotor mounted to part of the test apparatus.
The fabricated rotor mounted to part of the test apparatus.

The rotor was constructed on a soft steel (A36) annulus, 0.25" thick and 2.2" wide. The (32) N42 Neodymium Iron Boron magnets, each 0.5" X 1.0" X 2.0", were placed on the the annulus and epoxy was used to glue the magnets in place. The magnets are extremely powerful and can be quite dangerous to handle. Caution is recommended.

[edit] Material Cost

Magnets $120
Wire (6 lbs.) $50
A36 Steel Ring $30
Miscellaneous $30
Total $230

[edit] Testing

Measuring the torque of the motor.
Measuring the torque of the motor.

A turntable using a large Inca bearing was fabricated to test the mounted motor. Torque was applied to the turntable using a calibrated spring, as shown in the figure to the right. The current to the winding was increased until the torque from the spring was balanced at the 90 degree electrical phase position.

The single-phase torque constant was measured several times for each phase. In every case the torque constant was determined to be 11 +/-0.25 N-m/amp, a factor of approximately two larger than predicted (see Table 2). The discrepancy was found to be in the simulated average magnetic field strength, which in practice is close to 0.6 Tesla for the selected magnets and geometry.

We note in passing that the motor described herein would make an excellent generator for a low speed wind turbine.

Dan Grays Testing Results: Dan Grays Test

[edit] References

Direct Drive Motor Powerpoint Slides by Dave Rowe
Motor Calculation Spreadsheet by Dave Rowe

[edit] Supplies

Magnets
Magnets and Wire
Personal tools