Those of you who are reading my post know that there is not much 'black magic' behind EMC and that there is always an explanation for why we fail during the EMC testing process. As a result of having a strong understanding of EMC, it is extremely satisfying to be able to pass the EMC testing for both yourself and the customer at the same time. In one instance, the testing engineer spent at least 30 minutes verifying that the configuration was correct before he sent it off to be tested before he sent it off for testing. In my question, I asked him: "Why are you checking?" and he replied: "The last time was awful, so I want to do a sanity check to make sure I'm not making any errors this time". This was a very pleasant surprise to me, and I take it as a compliment. The unit performed so well during the immunity test that the testing engineer was convinced he had made a mistake because it responded so well. To return to my point, I want to discuss here a design technique you need to master in order to prevent your diodes and operational amplifier from working as AM demodulators. Here's a simple diode rectifier to get you started. As shown in the post figure, there are times when we need to rectify a signal, and the easiest way is to use a diode. We have seen this kind of circuit many times during our university studies, right? This circuit creates an AM demodulator by doing that. You can see the demodulator if the output load of your diode is high, and if you imagine a 20 pF capacitor between the cathode and the GND. How about the modulated signal? In Conductive immunity testing (depending on the standard), the carrier is modulated in amplitude by a sinewave (usually 1 kHz). Whether it's conductive immunity or radiated immunity, this sinewave is present in both of them. Suppose you are performing either of these tests and you fail one of them. It is recommended that you try without modulation, and if the problem disappears, then you know that you have an unwanted demodulator in your design, and this post may assist you in solving it. When the AM demodulation is made with a diode, it is easiest to fix it by adding a capacitor (between 20 pF and 200 pF) in parallel to the diode, which will fix the problem. If you look at the drawing of the post, you will be able to understand what the capacitor is doing. When it comes to an operational amplifier, if you have demonstrated that an operational amplifier fails during its radiated immunity or conductive immunity, then why do you not use the same technique to prove that it fails during its radiated immunity? It is a good idea to add a small capacitor between the pins - and + on your operational amplifier. However, you have to be cautious here; not only are you adding positive feedback to the system, but you might also start oscillating the operational amplifier if the capacitor value you choose is too high. You will be fine if you have a phase margin of 40 degrees, so try to keep your phase margin low. It is my hope that you have found these tips to be useful. In response to a previous post, I have been asked to discuss the effect of vias when changing layers. I will talk about this in my next post, as we will see that there is not much to worry about for most cases.
Comments