Issue link: https://resources.pcb.cadence.com/i/1533020
Introduction to PCB Constraints I remember my very first PCB designs way back around 1990. We didn't even use a netlist, let alone a set of embedded design rules. In fact, the tool we used to create the RF amplifiers for the military and the industry was better suited to designing the floorplan of a house than an electronic circuit. Those printed circuit boards were strictly single layer boards with an unbroken ground plane on the bottom. They were subse- quently joined with a so-called pallet that was made of aluminum and bonded to the metalized bottom side of the board. We used the same software to design the circuit and the aluminum underpinning. Why Constraints Matter Thus, it was a big deal to finally get to use actual ECAD software to do the multi-layer control boards for the amplifiers. It wasn't as big of a leap forward as you'd expect, being the early '90s and all. In particular, analog circuits will feature some traces that are controlled by their width and also by the air gap to the ground plane. That's not asking for much. Figure 1. A test mule for various phone features. Size really matters in this space so we need tools that can adapt to the density. Image Credit: Author Having multiple air gaps was not supported with mid-90s "shrinkwrap" software. What I had to do was temporarily make all of the high frequency traces extra wide and create a ground plane around that geometry. Then, I had to change the RF lines back to their specified width to make Gerber data. Finally we used a Gerber editing tool to remove and replace the original shapes with the ones that were doctored for the novel spacing requirements. Mixing different Gerber layers to get something in between the two is not for those who are risk averse! 2 www.cadence.com Introduction to PCB Constraints