Reading time ( words)
I recently attended IMS 2015, a show that focuses primarily on RF and microwave technology. During the show, I met with Per Viklund, the director of IC packaging and RF product lines at Mentor Graphics, and Alex Caravajal, business development manager with Mentor Graphics. We talked about the challenges facing PCB designers working with RF and microwave technology, and Mentor’s efforts to help reduce the RF design cycle time.
Barry Matties: Per, please begin by giving a quick overview of what Mentor Graphics is featuring here at IMS.
Per Viklund: Specifically for this show, we're focusing on the RF. Our customers have told us for a long time that it takes too long to develop RF circuits, especially when RF is going to be a piece of a bigger system. It's not so often that RF engineers have told us that. It's the PC board integrators who receive the RF circuit for integration and the management who sees long engineering times, lots of re-spins, that clearly says it can't be real to have these long cycles, or it can't be correct to have to do five or six iterations. There's got to be a different way. Not often will the entire engineering organization admit that there is a better way.
Matties: When we talk about cycles, what sort of cycle times are we talking about?
Viklund: It could be different depending on if you're doing a simple project or a larger one. But we have some cycle times that can be over two months long, say, if you are doing a board for a satellite communication link or something like that. They could have very long cycles.
Matties: As you were saying, there are many iterations.
Viklund: There are many iterations, because first you try to simulate, and it doesn't work so you try again. And that intricate process takes a long time because every time you have to send data back and forth between RF design tools and board tools and integrate and simulate and so on. It just takes a long time. You might be surprised how many people still use the ASCII files that transfer RF circuits from, for example, ADS or National Instruments Microwave Office into board tools. And when they do it that way, it becomes a dumb metal blob that the tool cannot make anything smart out of. It's just a blob; you can place it and it comes out in your fabrication data. But the problem is, every time something changes, they have to do that entire process again because there is no intelligence.
So what we try to help these companies with is to say, "Well, if you have intelligent data in your ADS or Microwave Office tool, let's try to maintain that intelligent data in a round trip.” So that you can send that data into our tools and implement it into a bigger system, bigger context, but still maintain all the parametric shapes and programmable shapes that you have. You’re able to set up things like tuning expressions and then, by just a push button, send the data back into ADS or Microwave Office again to analyze and make adjustments, optimize and then just send the adjustments back into our tools again. But that loop has actually proven to shorten this design cycle by more than 50% in most cases. It’s significant.
Matties: People must be pretty excited about that.
Viklund: They are very excited about it. But it's hard sometimes to adapt a new methodology. If you've been doing one thing for 20 years and now suddenly we tell you to do something different, and you won’t see the benefit until the middle or end of your project rather than immediately, that's challenging to people. They say, “I know I have a pretty good project. You're telling me to use a little bit of time up front.” Yes, I am, and you're going to gain it back. People are more and more adapting to this.
To read this entire article, which appeared in the August 2015 issue of The PCB Design Magazine, click here.