RFI
[Top] [All Lists]

Re: [RFI] Power Over Ethernet from Switch Mode Power supplies

To: Chuck Gooden <Chuck.Gooden@comcast.net>
Subject: Re: [RFI] Power Over Ethernet from Switch Mode Power supplies
From: Greg Troxel <gdt@lexort.com>
Date: Mon, 08 Jul 2019 12:32:57 -0400
List-post: <mailto:rfi@contesting.com>
Chuck Gooden <Chuck.Gooden@comcast.net> writes:

> I am currently experiencing RFI issues on 40 meters and below. Some,
> but not all, of these issues are due to the Power over ethernet (POE)
> switches that I am using.  I have two POE switches and 2 Wirelass
> Access Points (WAP).  The switches are powered by a 48 volt 1.3 amp
> switching power supply and can supply power to 4 of the 8 ports.  I
> currently only use the POW on the switch to power a network cloud key
> controller.  The WAP's also have a 48 V switching power supply
> connected to an ethernet power inserted. So in this configuration I am
> using 3 of the switching power supplys.
>
> I am considering redoing my network wiring so I connect the two WAP's
> to one of the switches (#1) and power my second switch from one of the
> POE ports also on the switch.  This will involve running more ethernet
> CAT5 cable but will get rid of 3 SMP power supplies.

You didn't mention what kind of equipment, and I'm not sure why not, as
the collection of anecdotes is somewhat helpful to others.

I have measured some Ubiquiti access points (they come with 24V
supplies/injectors) and a Ubiquiti ToughSwitch (also 24V SMPS) that can
power the access points via POE.  I have friends who do EMC
professionally, and was able to borrow some equipment and make
measurements of noise conducted onto the power lines with the equipment
operating (and the switch actually powering an AP), with a legit LISN
and spectrum analyzer, and found the levels to be quite low (measurable,
but in a cluster with the better SMPS out there, in stark contrast to
the real junk which is 20-40 dB worse, speaking very roughly from
memory).  Of course, that doesn't cover noise on the ethernet wiring,
and probably some other things, and test measurements do not exactly
equate to received noise - but I find it usually correlates pretty well.

Overall, the Ubiquiti equipment seemed quite clean for modern stuff.

So, I wonder if you are using Ubiquiti (but you say 48, not 24, and
there are of course many POE standards from which to choose!), or
soemthing else.

Besides Jim's advice -- which you should follow seriously, and read his
publications multiple times through -- you might see if you can borrow a
linear 48V supply from a fellow ham and try it out.

I would also wonder if the access points or switch conducted noise into
the PS wiring or ethernet-as-power-wiring even if the supply were
clean.  Surely the access point does not actually use 48V and is
probably dc-dc converting it to 5 or 3.3 or something.

73 de n1dam
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
<Prev in Thread] Current Thread [Next in Thread>