[RFI] DC Power supplies

Doug Powell dougp01 at gmail.com
Wed May 4 19:07:52 EDT 2016

You are absolutely correct.

Although it is not possible to be dead quiet, with enough filtering,
shielding, and RF absorbers it is possible to reduce your RF signature to
extremely low levels. I have worked with an ferrite lined shield room that
needed a 120VAC 30 Amp service on the interior.  In order for the power
lines to penetrate the wall of that room we had a line filter that must
have weighed at least 50 pounds.  The MIL-STD-285 shielding effectiveness
of that room was a minimum 130 dB all the way up to 3 GHz.

In the last 10-15 years a lot of effort has gone into the design of
resonant and quasi-resonant switchers, these can be very low noise.  To be
really honest, I haven't done a lot with switchers below 1,000 watts lately
so it's entirely possible some new technologies are available.

In case you're interested, this is an example of an emissions profile in a
standardized test setup. https://goo.gl/09xZ9J

Best, -Doug

Douglas E Powell
Laporte, Colorado USA
dougp01 at gmail.com

On Wed, May 4, 2016 at 4:16 PM, Charlie at thegallos.com <charlie at thegallos.com
> wrote:

> Actually, mumble mumble years ago, I worked on switchers that had, for all
> intents no measurable noise in both the RE and CE range of test from
> Mil-810. Of course the USN was spending major money for rack steel cased,
> mu metal wrapped, hand tweaked power supply we shipped. Spend enough and
> you would be surprised how quiet a switcher you can make
> 73 de KG2V
> > On May 4, 2016, at 5:59 PM, Doug Powell <dougp01 at gmail.com> wrote:
> >
> > There is no such thing as a "noise free"‎ switcher, only unaffected
> equipment.  The quietest switching power supplies tend to be medical
> grade.   You don't want RFI affecting life support equipment.
> >
> > All switchers pull pulses of current from the rectified AC Line at
> frequencies of a few kHz up to 1 Mhz. While this fundamental switching
> frequency and mostly odd harmonics can be part of the problem, the real
> issue is the reverse recovery time of the diodes and the storage time of
> the switching transistors. These times can be from a few microseconds down
> to 100 nanoseconds or less. The reciprocal of these recovery times is basis
> of the frequencies and the harmonics  they produce.  ‎Switch-mode power
> supply designers are always pressured for higher and higher operating
> frequencies and faster recovery times in order to shrink physical size and
> improve efficiency.   Just think about to latest generation of USB 3.0
> power packs and how small they are for the amount of power they produce.
> >
> >  The high frequency pulses produced by bipolar transistor designs can
> easily have a fourier content in the 150 kHz to 30 MHz range. The MOSFET
> type switchers can typically produce additional frequencies from 10 MHz on
> up to more than 300 MHz. The stuff in the low end tends to be narrow band
> emissions and the stuff on the high end tends to be broadband noise, for
> the most part.  Standard emissions testing these days set limits in a range
> from a few 10s of kHz to 3 GHz and sometimes higher.
> >
> > So the bottom line is to find a power pack with a noise profile that
> works in your application.  If it is of interest to the group, I may be
> able to pull up some old plots from a spectrum analyzer, just to show what
> I mean by "noise profile".
> >
> > All the best, ~ Doug
> >
> >
> >   Original Message
> > From: WD8ARZ
> > Sent: Wednesday, May 4, 2016 3:15 PM
> > To: rfi at contesting.com
> > Subject: Re: [RFI] DC Power supplies
> >
> > AC switching type power supplies tend to be very rfi noisy. Transformer
> > type power supplies dont use the the dc to ac to dc switching that
> > generate that RFI .... and should be quieter power supplies. They are
> > heavier, with line isolation in most cases, and worth the extra expense
> > and weight. Another bad trait for those switching power supplies is that
> > there is generally not a true physical break the utility power feed
> > switch. Instead they just disable the output and leave the switching
> > circuits active and thus are RFI sources as long as they are plugged in.
> >
> > 73 from Bill - WD8ARZ
> > South Bend, Indiana
> >
> >> On 5/4/2016 2:51 PM, charlie at thegallos.com wrote:
> >> Hey Gang,
> >> We all know about buying old analog wall warts for power supplies to cut
> >> down on RFI, but sometimes, that is NOT going to happen
> >>
> >> snip snip
> >>
> >> I think it might be a real good idea if we could come up with a list of
> >> various "Line lump" (so they can be wired in) type power supplies at
> >> various power ratings that are KNOWN to be good, and sources for them
> >>
> >> I'd be more than happy to collect the list and put it up on my web site
> >>
> >> 73 de KG2V - Charlie - www.thegallos.com
> > _______________________________________________
> > RFI mailing list
> > RFI at contesting.com
> > http://lists.contesting.com/mailman/listinfo/rfi
> > _______________________________________________
> > RFI mailing list
> > RFI at contesting.com
> > http://lists.contesting.com/mailman/listinfo/rfi


Douglas E Powell

dougp01 at gmail.com

More information about the RFI mailing list