At the 2007 Minos micromouse conference, Martin Barratt gave a talk about his stepper motor driving scheme. For best performance, steppers really need to be driven by a high voltage. However, the current will need to be limited to prevent the motor overheating. Normally this is done with dropper resistors which waste all the energy or with chopper drives that can be relatively complex. Martin has an alternative approach that provides the high voltage needed to produce fast step rates and the low current needed to reduce dissipation and keep the battery size under control.
The essence of the method is quite simple. Stepper motors look like inductors to the driver circuit. The torque available depends upon the current flow through the coils but the rate at which the current can increase is proportional to the applied voltage. If you drive a stepper at low voltage and low frequency, it is perfectly happy. You will get adequate torque as the rise-time of the current will be a small part of the total step time. As you increase the step rate, there is less and less time for the current to rise and the torque suffers greatly. Simply using a higher voltage will result in the coils being over-driven severely. This will require a large battery and will make the motor overheat significantly.
Martin’s solution is to have a two level power supply and a switch. At low speeds, or when the stepper is holding, the coils are supplied from a lower voltage. When a step is made, the switch connects a high voltage to the motors for a short period – long enough to get the current up to a respectable value quickly enough to ensure the motor responds quickly.
You can see the entire presentation with a detailed description of the rationale and the implementation here:
the proof of the pudding is in the eating and Martin’s mouse uses steppers motors and this drive technique to achieve excellent performance with only a modest set of batteries.