Motor controllers: build or buy?

At maxon motor, we often hear experienced engineers' talk of developing their own controller for their application, as it's perceived as being an easy thing to do and, therefore, a cheaper option. But is it?
If you haven't developed a controller before, and are considering whether to build or buy, here are some of the factors to be aware of.

The solution and the effort for optimising the control algorithms, the power stage design and current measurement is very often underestimated. This task is made more difficult when trying to control high dynamic, low inductance, coreless/slotless permanent magnet motors. Sensorless control of brushless (electrically commutated) motors is even more demanding and the success strongly depends on the application's requirements. The effort of developing a controller and power stage from scratch (based on a general application note) will "cost" at least 6-12 months of development time and the “risks” involved will require management and liability insurance. The developed controller will also have to be inspected, and maybe certified (e.g. EMC, electrical safety, installation) to the standards required by the application.

Engineering a controller to run a specialist motor is not just a simple matter of selecting a microcontroller and copying the schematics of an application note, it is considerably more complex and there are often unforeseen issues that can jeopardise the success of the entire project. These complexities increase with the demands of the particular application and its environmental conditions; highly sophisticated sensorless control for low speed or changing load operation, the task gets extensively more demanding and would take 12-24 months of development costs.

maxon can offer customised controllers with fair development costs and a reduced development time due to our wealth of experience with different DSPs, control algorithms and power stage design. We do not have to start from scratch and we do not have to rely on application notes and best effort guess work. Our knowledge is based on more than 20 years of practical controller design in a wealth of different applications and environments. The testing and certification can be facilitated and because of our experience the units are more likely to pass these certifications not wasting precious development time on multiple revisions and repeating attempts to achieve those specified certification levels.

In summary:

• Unless you have specialist knowledge of servo motors and controller design, try to buy off the shelf or a modification to a standard product.
• If your specification requests that the intellectual property of the device; firmware and inner workings is known by the customer you may want to consider making your own controller.
• The long term costs of designing, building and certifying a DIY controller are often overlooked and can cause projects to become delayed, over budget and not perform to expectations.
• Why reinvent the wheel?

If you would like a customised product please contact us with your requirements.