当我到LA参加EDN编辑会议时，见到他们，才知道Semtech是个好公司。不仅如此，来自于Building T的伙伴Craig，在硅谷辛勤的为他们工作。Semtech正在做越来越多的模拟产品，如模拟开关，场效应管驱动和电荷泵等，但他们也做些简单的保护电路。因为像STM公司制造的滤波器，设计理念很简单，以至于新闻界往往忽视它们。如今制作一个二极管保护装置，看起来似乎微不足道，但像ST公司用很小的杂散电容(0.6pF)，或像Semtech公司EClamp™238xp产品中结合RC元件来做，即使不涉及任何复杂模拟电路，那也是真的高科技了。这是过程工程学、物理学和制造业的胜利，而不是电路设计。Semtech公司为EClamp2384，EClamp2388系列给出了发行版本，最后一位数字代表保护了多少线路。
保护电路避免射频干扰的问题越来越严重。当我在National S工作时，我曾采访过一个局域网VOIP电话制造商。离办公室电话一英尺远的手机信号干扰了音频电路。导致电路板重新设计，在关键节点处增设小电容，以保持手机射频远离模拟音频信号路径。而工程师犯的另一个错误是使用大容量、大体积电容器。你可以试着在900、1800或1900MHz处滤掉射频。在微小的0402或更小的情况下，需要皮法，为的是在这些频率处有低阻抗——不管你打算采用哪种电容，肯定得看看阻抗情况。更好的选择是，使用Semtech公司的元件，剩下的只需要元件能确保射频干扰信号远离电路。
我正在设法滤除来自于办公室电话的射频干扰。如果你在设计一个手机或，干扰甚至是很严重的，因为间距是英寸而不是英尺。你的产品可能是蓝牙，802.11以及相机和手机系统。保护电路不受各种干扰源的射频影响是困难的。像大多数EMI问题一样，在你想要定型之前，它就冒出来了，所以要确保明确这个问题；这个问题正越来越频繁的发生。
英文原文：
Semtech R-C filters for EMI in handhelds
I know Semtech are good people from meeting them when I was down in LA for an EDN editor’s conference. Not only that, my buddy Earl Craig from Building T works for them up hare in Silicon Valley. Semtech is doing more and more analog stuff like switching regulators. FET drivers and charge pumps but they also make some simple protection circuits that can save you a lot of grief. Because these filters, like ones STM makes, are so simple in concept, the press often overlooks them. Now it may seem pretty trivial to make a diode protection device, but to do it with really low stray capacitance (0.6pF) like ST does, or to combine R-C components like Semtech does in their EClamp™238xP, well that is really pretty high-tech, even if it does not involve any complex analog circuitry. It is a triumph of process engineering, physics and manufacturing, not circuit design. Semtech sent the press release for the Eclamp2384, Eclamp2384, Eclamp2388 (pdf) series of parts, the last digit indicating how many lines it protects. Protecting your circuits from RF interference is getting to be a big deal. I called on a LAN VOIP phone maker when I worked for National S. They had interference getting into their audio circuit from a cell phone set down a foot away from their office phone. It took some board redesign and the addition of small capacitors on critical nodes in order to keep the cell phone RF out of the analog audio signal path. And one mistake engineers make is using big-value physically-big capacitors. You are trying to filter out RF at 900, 1,800 or 1,900 MHz. You need picofarads in a tiny 0402 or smaller case in order to have low impedance at those frequencies—be sure to look at the impedance plots for whatever cap you are thinking about using. Better yet, use parts like Semtech makes and rest assured that the parts will keep interfering RF signals out of your circuits.
I was trying to filter out RF form an office phone, if you are designing a cell phone or handheld it is even tougher because then you have inches, not feet of separation. Your product might have Bluetooth, 802.11 as well as camera and cell phone systems. Keeping the RF from all those interfering sources out of your circuitry can be tough. Like most EMI issues, this one comes up right before you want to ship, so be sure to have an understanding of these issues; they are coming up more and more often.
英文原文地址：
