Duane Benson, I thank you for this article and you had some very good information to share. But I think you are still on the dark side. Engineer marketing? I am sure the people at Harvard Business School will laugh.
David - With a thru-hole part you want the capillary action to wick solder down through the hole to the other side. That same capillary action will cause the solder to creep further down the part leg causing the required fillet.
empty vias don't have the component lead to keep the solder in place and, especially with surface mount parts, everything tends to be smaller, so there's really no good place for the solder to go.
Duane, very interesting, thanks. But what is the difference (apart from maybe size) between C in your diagram (which you say is a NO NO NO!) and a through plated hole for a normal wired or DIP component?
Karen - I pass out this sort of information in a company blog. Probably nine out of ten posts on the blog are tips and hints for PCB designers - not too much marketing glurge. We also have a few white papers that are essentially the top ten list of things not to do.
Based on what I've seen and heard, there are a couple of different reasons that errors like open via in pad are so common. One is that fewer young engineers have older engineer mentors available to them. Another is that, with staff reductions, many companies that used to have PCB layout specialists on staff don't any more and the design engineers are now tasked with the layout, which isn't necessarily their area of expertise.
A long time ago, I worked for In Focus, the projector company. We had a layout specialist - Tom, who's last name I can't remember. This guy was an absolute wizard. He could be in the CAD layout software hitting key-codes with one hand, moving the mouse with the other, all while holding a conversation with someone behind him. Most engineering departments used to have that but very many have downsized that position away.
But wait - there's more... those same engineers are often tasked with more to do in a shorter period of time than was the case years ago. Sometimes the designer simply isn't allocated enough time to take the care to learn and avoid such mistakes.
Hi Duane, greast explanation of a common problem and why it's not a good idea. Why is it, do you think, that engineers continue to keep making the same mistake? I think back to my early days as an engineer, when the "old guys" would take younger engineers under their wing and tell us about all the bugaboos to avoid that they'd only learned through hard experience. Does your company, for example, include this in a "Ten Things NOT to do when designing a PCB?"
Tom - Thanks for the note. I'm glad you found part of the blog interesting. I'm now done with my introduction here and my inferiority complex has said its "justification" piece so I now longer feel compelled to pollute actual information with non-electronics type stuff.
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...