This article also appears in
Subscribe now »

SCALEXIO emulates electric motors together with the Electronic Load Module and DS2655 FPGA base module.

dSPACE enhances environment for electric motor controller testing

dSPACE’s latest solution for electric motor controller testing combines SCALEXIO hardware-in-the-loop (HIL) simulation technology with a DS2655 FPGA base module and a scalable electronic load system to enable users to test electric motor control systems in a more realistic environment by emulating the real motor and generator currents.

To be demonstrated in booth 303 at the SAE 2014 World Congress in Detroit Apil 8-10, the SCALEXIO DS2655 FPGA base module is user-programmable to define high-fidelity, custom models for simulating complex, highly dynamic systems. Testing of ECUs for electric motors typically requires short simulation cycles, fast calculations, and quick I/O access. Modules can be added to the DS2655 base board to provide the necessary number of I/O channels.

The FPGA device can be configured graphically with standard dSPACE tools, so that users can react flexibly to new requirements, such as new I/O interfaces or faster simulation. dSPACE also provides the simulation models for HIL tests of electric motors and associated power electronics and batteries.

“What makes it unique is that it’s a completely graphically programmable solution for motor control and motor simulation,” said Jace Allen, Lead Technical Specialist–Simulation & Test Systems, dSPACE Inc. “We have a lot of core libraries or functions that are designed to make using the FPGA extremely easy.”

The included scalable dSPACE electronic load module emulates motor and generator currents of up to 100 A at voltages of up to 60 V. This voltage range allows emulating vehicle electrical systems of up to 48 V. The electronic loads can also be used in parallel to provide higher current capability.

“There’s really three ways to test an electric motor system or electric motor ECU,” Allen said. “One is to put it on a dyno and drive a real motor. The other way is to test completely at the signal level, where you bypass all of the power stages and connect the ECU to a test system and test everything at a low power signal level. Then you can also test with the actual power in the loop, and that’s what this electronic load solution provides. It’s a seamless way to put into a simulator the capability to dynamically simulate the loading that an inverter would experience as if it was connected to a true motor.”

The Electronic Load Module features energy-recovery capabilities, increasing overall system efficiency and providing energy saving during testing. As a result of the lower energy consumption, less heat is generated, therefore less active cooling is necessary.

“This uses an inverse power energy-recovery methodology the way it runs, so that it doesn't dissipate all of the power as heat,” Allen said. “We reduce the power with an inverse power feedback in a sense. So then you don’t have to deal with as much heat dissipation; it makes it easier to put together a test system and it provides energy savings.”

The Electronic Load Module can be integrated within the SCALEXIO system or installed externally. The ECU under test can be connected directly to the Electronic Load Module, resulting in a short power signal connection and reduced possible signal interference. This approach also provides a method to test sensorless brushless dc motor controls by simulating their characteristics and possible variations.

This approach can pay dividends by accurately modeling nonlinearities in the simulation of these motors.

“Everybody is trying to package their motor in the smallest form factor for whatever their application requires,” Allen said. “When they do that, they can introduce nonlinearities in the motor characteristics and how it performs. So you try to build a controller that compensates for that and optimizes the performance of that motor even though it has these nonlinearities. It can be very difficult to simulate those nonlinearities, and using an approach like this we can interact with tools like JMAG to allow users to virtually model any type of a motor and simulate it in real time.”

Continue reading »