Get email delivery of the Cadence blog featured here
I was reading an autobiographical narrative by Kenneth Anderson (1910 – 74), a renowned Indian wildlife expert of Scottish descent. Now, Mr. Anderson was also an automobile enthusiast – loved toying around with his fleet of second hand Studebakers. He describes how he ripped apart a Studebaker, tweaked the engine, and reassembled a lighter version resembling an oversize cart with a cane chair as the driver’s seat.
But that was in the early twentieth century and, if you don’t count radios, a car was a mechanical machine. Today, the automotive industry is no more producing a purely physical vehicle: it is an interplay of physical and electronics components. Starting with a few or even a single processor in the latter half of the twentieth century, today the automotive industry is grappling with scores of processors for power control, traction systems, safety systems, and transmission control; not to mention the non system-level applications like infotainment, air conditioning, and navigation. Automotive embedded systems with dozens of electronic control units (ECU) have become the norm for cars.
Alas, Mr. Anderson would not have been able to change any of today’s cars that easily, and messing with the engine? Almost impossible for a single man with a few mechanical tools. But why only cars? With IoT, you already see many of the ubiquitous household devices moving in the same direction; electronics and mechanical parts interacting with each other.
But you are most probably not worried about re-assembling a car or tweaking engines for fun; you make these products – the cars, the IoT‑enabled consumer devices. You design them and manufacture them. Worse, you design at the system level. Your worry is doubled, or quadrupled even. So, how do you ensure your design works? You simulated the mechanical parts, but what about the electronics? You need to simulate and model analog/digital mix-signal electronics alongside mechanical, hydraulic and thermal parts. You are worried about escalating cost and time-to-market.
You might have dabbled with prototyping as a solution: used a behavioral simulator with ideal mechanical parts, or used a circuit simulator with ideal electronics models. But ideal models give ideal results; and, if you used a simulator built for mechanical or physical parts, the results are almost as good as ignoring the electronics. When you move from prototyping to real production, you need to fine tune again because ideal parts cannot emulate the realities of delays and non-linearities. This costs money and eats into production time, affecting your bottom-line and time-to-market. Sometimes the worst happens; you are forced to recall your products, say, because your claim in good faith based on simulations turned out to be far from the expected, in reality.
What you need is to be able to identify design issues earlier in the design phase, that too, with real models.
If you are nodding your head while reading this, most probably PSpice® Systems Option will interest you. The biggest advantage of PSpice Systems Option is that you use a single prototype to co-simulate electrical and mechanical systems. It uses the PSpice® Simulink Co-Simulation Interface to substitute electronic blocks in PSpice, while the rest of the design is simulated using Simulink.
To know more, you can attend this Webinar on combining MATLAB and Simulink with PSpice to streamline PCB designs. Or, watch the videos, hear the experts, and generally explore the technology.
Good luck exploring PSpice® Systems Option while I explore the jungles of yore with Mr. Anderson!