The Antenna Guy

Hello folks.  There's been some time since I've written my last post. I've been very busy with work and other side projects, which eventually will find their way here, but since they're far from completion, I thought to make a smaller post about a subject that is useful and I've had had in my notes for a long time. So I figured I'd just put it up on the blog so it could be useful as reference for someone, or even myself one of these days. The subject is simple, how to tune a dual-band antenna. There's a fair amount of articles and tutorials, even on YouTube, about matching networks, either LC and Pi networks. They'll teach you all you need to know about Smith charts and what not. That's all very useful, but they all have one problem, they usually focus on having impedance \(Z_1=R_1+jX_1\) become \(Z_0\) \(\Omega\) at one single frequency. But how do you do it when your antenna is suppose to operate at two different frequency bands? How to tune both b...

A while ago I promised to look into the effects of using VIAs (Vertical Interconnect Access) on RF tracks to change layers on PCBs. Unlike the 90º degree bends on PCB tracks, in this case there's essentially two postures about the matter. Either people are completely clueless of the potential risk of placing a via on a RF track and do it recklessly. Or they're somehow a little aware of the risks, therefore completely demonize such practice, going at extreme lengths such as increasing the track sizes immensely (making it even worst), just so to avoid the use of the via.  A long time ago I found an awesome primer about vias for high-speed and RF tracks from Keysight, you can find it here . However, it focuses mainly on multi-layer board in excess of 8 layers. Therefore, although the concepts are the same and the material is absolutely valuable no matter your scenario, it lacks the actual results you'd get in a 2 or 4 layer board.  I've been cooking this post for a while, ...

Hey there folks. Today I'm going to talk about curves in PCB tracks. I've seen a few topics here and there trying to demystify the way that tracks should curve in order to avoid all sorts of magic stuff. There seems to be a large consensus about avoiding 90º degrees bend at all costs, then there's a lot of people already claiming we've been doing stuff based on tradition only and that there's no solid reason nor practical argument to not use 90º bent tracks. So I thought I should give my two cents on this subject. I f you want to indulge in such a discussion, check out this EDA stack overflow thread . The most voted answer of said thread is certainly very detailed and mostly demystifies religious beliefs about the 90 degree angles in the tracks. But I also feel it ends up going overboard with statements such as: "In fact, this is true even at very high frequencies, which I will address further down in this post. If I...