[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] Corcom EMI filter



Gary Lau wrote:
The filter case should be connected to the mains green-wire ground.  I've
written a web page on this very non-intuitive topic:
http://www.laushaus.com/tesla/emifilter.htm



I generally agree with Gary's observations on the linked page..

What's interesting is that these filters are designed not to filter out interference coming from the outside, but to reduce noise being propagated out from your circuit.. so the "use it backwards" is entirely the wrong way to use them. If you think about it, they get used because manufacturers want to meet a FCC or EU "conducted emissions" test, not because the mfr cares about transients from the outside world. So how the mfr intends it to be used and how we intend it to be used in filtering the supply for a TC is exactly the same.



A couple things to watch out for..

If you add a capacitor between line and greenwire ground (on the input) it has to be appropriately rated. If it fails because of a line transient, it will sometimes fail semi-shorted, and just draw a lot of leakage current, get hot, and that is bad. (10mA at 120V is a watt, and a little disk ceramic gets mighty hot at a watt)

You may not want a capacitor to green wire ground on the input anyway.. A cap that is big enough to shunt 100kHz interference to ground is probably big enough to draw enough current to trip a GFCI. The leakage spec on the filter is probably around 1 mA.

The Corcom filter that Gary shows has the ground point on the load side, which is the typical scheme. (http://www.cor.com/) Unfortunately, the F2411 is an obsolete type, so there's no datasheet on the site. But I found a "K series" RFI filter which looks similar (in terms of schematic), so I looked at a 30VK6

http://www.cor.com/pdf/K.pdf

That filter has an attenuation (in a 50 ohm system!!!) of only 10 dB at 100kHz for common mode, 20dB for differential mode. It really doesn't start to "work" until you get up around 1 MHz.


_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla