Development History of the Patented Variable Torque Motor - Solutions

Motor Design Criteria:

After preliminary designs and consulting with Jim Hendershot, known worldwide as an expert in motor design, the decision was made to use a brushless DC motor. The choice of brushless DC was necessary for Zepp and Hendershot to meet the following goals:

• Peak efficiency of 90%, which is 15-20 points above most PM motors.
• 40-50 hp per motor
• High torque in compact package
• No brushes
• Family of motors. By using a modular design to allow basic steel laminations and magnets to be stacked in varying lengths.

Critical Assumptions:

• Higher motor efficiency would be increasingly important to electric vehicle manufacturers.
• The cost of power electronics would reduce more than 30% in three years.
• The cost of neodymium magnets would reduce more than 30% in three years.
• The cost of fuel would increase more than 30% in three years, increasing the interest in electric vehicles.

Results:

All of the assumptions were realized during the next three years. However, two other problems arose that needed to be solved before commercialization:

• Back EMF creates limits on motor speed ·
• Need to increase efficiency over wider range of RPM's

These problems were solved. How?

Problem:

Back EMF. As higher torque PM brushless motors were designed, motor top speeds were restricted due "back EMF" or generated voltage. While our motor design provided excellent low-end torque, tow tractors could not achieve enough top speed when lightly loaded.

Solution:

A two speed automotive-like manual transmission would solve the problem except: airport baggage handler work rules specified that some type of automatic transmission was required...and an automatic transmission wasted approximately 40% of its energy before it locked up. Such waste unnecessarily drained energy from the battery.

When studying the physics of permanent magnet motors we discovered weakening the permanent magnet strength would extend motor top speed and act as a transmission. Voila, the solution, but not before many designs were tried.