Motion Is in the Cards at PMD

View Prodigy Motion Cards at a Glance!
In a bid to make distributed motion control systems easier to implement and more cost effective, Performance Motion Devices (PMD) has come out with a new line of motion control cards that run their own motion programs without the need for a host computer.

These Prodigy Motion Cards feature an onboard code-execution module that PMD has dubbed the "C-Motion Engine." The module stores and independently executes downloadable motion control programs created in PMD's C-Motion, an integrated development environment based on C/C++ programming. Aside from the motion control functions, the C-Motion Engine can also manage digital and analog I/O signals.

The cards come in three different versions. Two retain a bus connection – either PCI or PC/104. A third version is a stand-alone card. All three support Ethernet, CANbus and serial communications.

And thanks to those communications supports, all three can still talk to to host computer. Only now, the host computer's role is relegated to sending high-level commands to the motion cards and running ancillary software. "The idea here is to offload the motion processing from the host and put the real-time motion tasks on a real-time platform," says Chuck Lewin, the company's founder and vice president of engineering.

The cards enable distributed motion control architectures that address some "classic motion control problems," says Lewin.

One is how to handle the motion on machines consisting of multiple stations or cells. With a distributed architecture based on standalone Prodigy motion cards, each cell can run its own motion program, while a supervisory computer sends the commands that coordinate the actions of the cells. "It's a good way to divvy up a big, complex motion problem into manageable pieces," Lewin says.

Another is how to sync motion systems with the peripherals they move – such as print heads or optical devices.  Lewin says systems with locally executed motion code and I/O management can often achieve a faster overall response than systems whose motion code runs on host computers that devote some of their processing power to non-real-time tasks.

While the standalone Prodigy card is one obvious choice for distributed applications, the bus-connected card can fit into distributed architectures too. According to Lewin, the bus connections offer the ability to create master-slave configurations consisting of multiple motion cards or to integrate cards physically into industrial computers.

PMD's take on distributed motion architectures will likely have cost implications. The company is selling the cards for $575. And Lewin says the card price doesn't tell the whole cost story. "There are obvious savings when you replace computers and I/O cards with a single motion card," he says.

Like the company's previous cards, Prodigy's motion processing muscle still comes from PMD's Magellan Motion Processor, which has a servo loop rate of 50 µs/axis. It offers a variety of motion modes, including S-curve, trapezoidal, velocity contouring and electronic gearing, while its servo loop compensation utilizes a full 32-bit position error, PID with velocity and acceleration feed forward, integration limit and dual biquad filters.