On 9/15/14, 10:02 AM, John Lloyd wrote:
Jim,
Many of today's switching power supplies with RFI filters in the AC
input have these capacitor from the line input to ground.
The pi filters on the input?
Yes, but the capacitors are also pretty small, designed to keep the 10s
of kHz switching currents from propagating back up the line, so the
leakage current is pretty small.
They also make sure that they have sufficient ratings, etc., to not
create a hazard.
For example
http://www.vicorpower.com/documents/user_guides/brick/UG_PFMVIBrickFilterEval.pdf
shows a typical input filter.. it's 4700 pF after the bridge rectifier,
and with the balanced choke, I think the AC Line/Ground leakage will be
quite small.
I think most EMI filters are differential mode (e.g. between the two
line conductors), not "line to ground" (because of the issues with
safety, etc.)
That's very different from "two 0.1 discs soldered across the line cord"
,or, 0.01 uF as shown on page 123 of my 1972 handbook "A 700-Volt
General Purpose Supply for Transceivers"
Or the similar balanced Pi filter for RFI reduction shown in Figure 16-1
on page 483
John Lloyd, K7JL
Message: 6
Date: Mon, 15 Sep 2014 07:36:53 -0700
From: Jim Lux<jimlux@earthlink.net>
To:towertalk@contesting.com
Subject: Re: [TowerTalk] Grounding Multiple Buildings
Message-ID:<5416F985.6010304@earthlink.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 9/15/14, 7:09 AM, David Jordan wrote:
Here is how the NAVY would perform the grounding:
http://www.hnsa.org/doc/radio/index.htm
-
How the navy would do it *in 1946*.
And, there's actually nothing in there about grounding, or electrical
wiring practices. The antennas they talk about are dipoles, for the most
part.
In general, old documents aren't always a good source of "good
construction and engineering practice". There has been substantial
change in grounding, bonding, transient suppression knowledge and
practice in the last 70 years, particularly when it comes to safety.
This is not to say that information in old publications is actually
wrong, but that it might be incomplete or inappropriate in view of
today's practices. A nice example is old versions of the ARRL handbook
which recommend connecting 0.1 uF capacitors between power line and
chassis for filtering. Today, this would be considered very bad
practice: capacitor failure would lead to a potential line/chassis
short, leakage currents (0.1 uf = 5 mA @ 120V) that will trip a GFCI, etc.
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|