RadioBanter - View Single Post

Telamon · August 11th 06, 10:24 PM posted to rec.radio.shortwave

In article . com,
wrote:

Sometime last year I started a thread about the RF noise from wall
warts. http://home.computer.net/~pritch/shortwav.htm

Telamon and I discused the best values of the diode "bypass"
capacitor.

For those inclined to do the job "right", here is a link to a very
detailed method.

http://www.hagtech.com/pdf/snubber.pdf

Enjoy.

Since noise limits my reception, I try to reduce it to as low a value
as I can.

That is a nice paper on the subject of snubbing. You get the most
performance out of a few components with just a capacitor and resistor
in series across a power supply switching diode. This is done for each
diode in the supply. The value of the capacitor calculated to couple the
peak transient power at switch off to the resistor to burn or dampen.
You also have to consider the RC time constant relative to the spike
duration. The capacitor and resistor form a small current loop around
the diode so the switching transient does not propagate through the
connecting circuitry as common or differential mode noise.

Snubbers are used for two reasons:
1. The already mentioned common mode noise reduction. For the switching
power supplies I used to work on the goal here was to reduce the spike
noise on the outputs and the EMI out the power cord.

2. Protect the switching diodes against the reverse voltage spike at
turnoff. The spikes at switch off can easily go over the diode rating
and the snubber is used to lower the spike below that rating.

Since there is a price to be paid for everything here it is power
efficiency. Faster diodes are more efficient but cause larger switch off
spikes that require a snubber to burn more power. Optimizing is the fun
part. Everybody thinks power supply design is a piece of cake, well it
is not when you try for high performance.

--
Telamon
Ventura, California