Issue link: https://iconnect007.uberflip.com/i/1315894
26 DESIGN007 MAGAZINE I DECEMBER 2020 ficient DC current gain to drive the 5V relay. Ultimately, a respin was needed to address the issue. The cost was in the millions of dollars, and the delay was over six months. Most often, I see respins occur due to sim- ple design mistakes rather than some complex multiple-variable problem that exists. The les- son is that simulating your designs will reduce the risk of respins due to simple mistakes or bad assumptions. Peer reviews are great, but SME reviews are better. Early in my career, I fell victim to the fallacy of allowing only my peers to review my work. I was responsible for the redesign of a digital processor/IO control board in an air- borne surveillance system. The control signals to the RF transmitter were TTL (5V), and my task was to interface them to an LVTTL (3.3V) FPGA. Simple enough, I thought; I will use a 3.3V-to-5V translation buffer. My peers in the preliminary design review agreed that this was a simple solution that was implemented cor- rectly. But during system integration, an engi- neer noticed that when the system was pow- ered on, a high-powered RF pulse was being transmitted from the system. I did not understand what was happening and just happened to speak with the power supply expert on our staff. When he saw my schematics, he instantly discovered the cause of the problem. He pointed out to me that the 3.3V and 5V power supplies were not sequenced together in the design, and in fact, the 5V leads the 3.3V by several milliseconds. During the time when the 5V is up and the 3.3V is off (~0V), the 5V transmitter control signals from the translation buffer are com- manding the transmitter to transmit at full power. This resulted in a respin of the proces- sor/IO control and power supply boards in the system. Including our power supply expert in my design review would have avoided two board respins in this project. Lesson learned: Always include subject matter experts in your design reviews. They have valuable knowl- edge and insights that can help keep you out of trouble. Respins do not have to be a part of your company's reality. If you enjoy PCB design, invest your time in becoming good at your job. Being interested in PCB design gives you the tenacity needed when facing difficult or com- plex design problems, and it provides the fuel for improving your design skills. A litmus test for gauging your interest in PCB design is how you spend your free time. Do you find yourself daydreaming about PCB design in your off time and/or tinkering at home with your own designs? If the answer is yes, then I suggest you find others who enjoy it as well, including SMEs from across the PCB design spectrum. Interact- ing with those who share your interest in PCB design helps strengthen your own abilities. An excellent resource for connecting to others with strong interest in PCB design is the Printed Cir- cuit Engineering Association. PCEA's core val- ues of collaboration, inspiration, and educa- tion are derived from the founding members' expertise and interest in PCB design. My inspi- ration for sharing my ideas and experiences in this article is derived from my interest in PCB design and encouragement from like-minded colleagues and friends. Remember: A sim in time will save dimes. And always let the appropriate subject matter experts take part in your design reviews. You might save your company millions of dollars and months of lost time-to-market. DESIGN007 Chris Young is chief hardware engineer for The Goebel Company and owner and lead engineer with Young Engineering Services. The lesson is that simulating your designs will reduce the risk of respins due to simple mistakes or bad assumptions.