I spent several stimulating days last week performing EMC testing on one of my most challenging designs in the past two years. This is the first time EMC compliance has been achieved in a topology of medical products. For this particular design, the EMC house was my most critical testing location; I foresaw three crucial challenges. ESD, conductive emission, and radiated emission. I managed to do a fast transient and started the Surge Test on the second day as well because the tests performed well. Every evening, at about 9:00 pm, I got a call from my customer asking, "Francesco, how did it go?" and a call from my colleague in the US discussing the success of our design decisions, "Francesco, well done!" We were correct in adding the CM choke and the voltage suppressor", "Francesco, did our solution work on ESD?" etc... Surprisingly, we had no problems; the testing turned out so well that I asked the testing engineer to complete the remaining tests on my own. Is there a reason for my statement? The following is a typical mistake I've seen when using DC power wall adapters in our designs. Since a power supply is certified, we often assume we can use it without taking any precautions. This doesn't make sense, and I will show you how it can all turn out wrong. The DC wall-mounted power supply can be connected to our electronic equipment in two ways. 1. Our equipment should be connected to the DC wall adapter and then connected to the AC mains by the wall adapter. 2. It is also possible to connect the DC wall adapter to the AC primary first, and then connect it to our equipment later (this could potentially damage the equipment). Assume that our AC wall adapter is 15V DC and our front end has components rated 25V, for example. In our front end, we only use ceramic capacitors (a ceramic capacitor has a very low ESR). In such a case, I expect an overvoltage to be as high as 30V (double the DC voltage rated). This can be seen if you examine the lead cable as a series of 1-10 uHenry inductors connected to a switch and to our LESR capacitor. What can we do to prevent this from happening again?
A method that I usually use is the use of a TVS in parallel with the capacitor. Sometimes I use a TVS in parallel with a varistor to ensure protection during fast transients and surges. Alternatively, electrolytic capacitors or even 1-ohm resistors could be added to the larger capacitors instead of bulk ceramic capacitors. With LTSpice or equivalent, you will discover that only the bulk capacitors need to have a poor ESR, while the 10/100 nF capacitors can be ceramics. Thank you for reading, and I hope you have learned the importance of adding protection to the input front end.
Comments