Issue link: https://iconnect007.uberflip.com/i/1256432
JUNE 2020 I DESIGN007 MAGAZINE 15 Hartley: Besides grounding problems, and we both know there are a million grounding prob- lems in app notes, most of them have that wrong. What are some of the other common errors that you've seen? Beeker: They sometimes have an overly abun- dant desire for isolation in the power supply. They love inductors and ferrites, and they end up in the application notes; it's aimed at prod- ucts, but they are coming from a board where they were going to do an evaluation for silicon. They have to put all these special hooks in there to isolate each pin so they can do things to it, but they don't remember which things were there for testing, and which things were there to make it work. They throw all that stuff in there, increasing the bill of materials cost and making the design far more complex than it needs to be, costing extra layer pairs and of- ten not working. Hartley: And costing extra, and with extra com- ponents, too. Beeker: All of that makes it much more diffi- cult to do a good job. The other thing is the ar- bitrary insistence on length matching. I see a lot of that, and it makes me crazy. They pick an arbitrary number that is admittedly arbi- trary by the memory vendors, for example. In our case, we will have an application note, telling you how to design your memory inter- face. They will verbatim copy from their favor- ite datasheet for the memory company they're working with and throw it in there with no consideration about what's required because it's a state machine and you're trying to meet the timing requirements. It's a lack of under- standing. Barry Matties: How do you know when and when not to follow an app note? Do you have to do all of this due diligence for every app note? Hartley: No. You have to be educated enough as an engineer to know what's good and what isn't good. When I went to that company in 2003, the one with the serious problems due to excessive length matching of memory lines, I already had almost 45 years of experience under my belt and was able to recognize the problem very quickly. Matties: When you think of the young engi- neers who don't have that, are coming out, and want to have a career in layout, what do they do? Hartley: That's the challenging part. I hate to say this, but it comes down to new engineers. If no one has told them that random layout ad- vice doesn't work, then they will have to find out the hard way. That's the bottom line. New engineers are going to have to work by trial and error and figure out what works and what doesn't. Dan hit the answer. Start reading and attend conferences. Beeker: This isn't taught in school, and they don't want to change what they're teaching. When I was first learning, I was confused. I'd go from Rick's class to somebody else's class, and with as much passion, they'd be saying Dan Beeker