RE: Why do TC's use line filters wired in reverse

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Malcolm Watts by way of Terry Fritz <twftesla-at-uswest-dot-net>" <m.j.watts-at-massey.ac.nz>

Hi Gary, all,

On 28 Feb 01, at 8:03, Tesla list wrote:

> Original poster: "Lau, Gary by way of Terry Fritz
> <twftesla-at-uswest-dot-net>" <Gary.Lau-at-compaq-dot-com>
> 
> After smoking a 30 Amp EMI filter, I looked at the filter circuit more
> carefully, and found that it DOES matter, a great deal, which way
> (line or load terminals to NST) it is hooked up.  It all depends upon
> what you tie the filter case terminal to.  If you tie the filter case
> to the 3rd wire AC ground, then you need to hook up the filter with
> LINE terminals to the NST, and LOAD terminals to the power source.  If
> you do like me (which I DON'T recommend) and tie the filter case to RF
> ground (and to the NST case, etc), the LOAD terminals should go the
> NST, and the LINE terminals should go to the power source.  The reason
> I connected the case to RF ground was that I didn't have a 3rd wire
> ground connection available at the filter.
> 
> If you don't connect the filter case to anything, then it probably
> doesn't matter which way you hook up the input/output, but I think the
> filter's effectiveness will be diminished.
> 
> The filter circuit is not symmetrical in-to-out, see the schematic. 
> There are very high voltage transients between a running coil's RF
> ground and the 3rd wire AC ground.  If the connections are not made as
> recommended in the previous paragraph, these transients will be
> applied directly across the filter's internal caps that connect to the
> case terminal (and possibly smoking them as I did).  And the filter
> will do a very poor job at keeping the transients out of your
> household wiring.  I have a page on my web site that details this. 
> See: http://people.ne.mediaone-dot-net/lau/tesla/emifilter.htm
> 
> Terry, do your AC EMI filters not have the case terminal
> asymmetrically connected as my schematic shows?
> 
> Gary Lau
> Waltham, MA USA
> 
> >Original poster: "Terry Fritz" <twftesla-at-uswest-dot-net>
> >
> >Hi All
> >
> >I deal with a lot of line filters at work.  All the ones I have ever
> >seen are exactly the same electrically going from "in to out" as
> >opposed to "out to in".  The terminals may be different on each end,
> >but they are electrically the same either way...
> >
> >I also checked a bunch of standard line filters I use here and they
> >are symetrical too.  The only TC filter I know of the is "out to in"
> >critical is my NST ouput filter (see how I am :-)) at:
> >
> >http://hot-streamer-dot-com/TeslaCoils/Misc/NSTFilt.jpg

My personal rule for hooking anything up in coiling is that the 
primary of the transformer should never be allowed to see more than 
mains potential between itself and anything else. That means mains 
filter ground terminals are connected to the mains earth as are 
transformer cases. The secondary side of the transformer is always 
left floating as far as the high voltage terminals are concerned and 
only the secondary and any discharge rods are connected to the RF 
ground. If I used filters on the secondary side of the transformer 
(which I don't), I would connect any ground terminals to the 
transformer case. None of the primary components are part of the 
Tesla secondary system and you can easily see for yourselves that RF 
ground wire going to the secondary can be elevated to considerable 
potentials due to the inductance and resistance in the RF ground path.

     On the question of symmetry in mains filters, it would depend on 
which you are using. Some are definitely asymmetrical from one side 
to the other. Consider a lot of popular common mode filters: you have 
a pair of chokes wound on a common core, a large cap hooked across 
one end and at the other, a pair of small caps, one from each line to 
a ground point (and sometimes a large one across both terminals on 
that side as well). Is it symmetrical with regard to high frequency 
AC? It depends on whether you are talking about common-mode or 
differential voltages I would have thought. 

Regards,
Malcolm