> Here's a thought about using capacitors in this way: We assume that
> the caps route the RF to a low-impedance load (the safety ground
> system) and that the efficiency of the driver (the TV set) will go
> down because its impedance is relatively high. The mismatch results
> in reduced radiation from the AC wiring in the house. This works if
> the impedance of the Safety Ground wiring is low. If it isn't, it
> could and probably will radiate as effectively as the hot and neutral
> wires did before the capacitors were added.
The primary enhancement of radiation occurs because the TV set (or SMPS in a
computer) has something to work "against". In computers, it's normally the
modem connection to the telco lines or other large wiring systems that act
like a counterpoise. Then the PS pumps the AC line, the AC line filter as
shipped from the factory is normally some rather large effective capacitors
from side to side of the line, but NOTHING to the chassis. So there is
substantial bypassing from conductor to conductor of the power line, but not
much if anythingto the chassis. The telephone line (in a computer) or the
antenna lead in a TV is bypassed to the chassis, so you have an instant
ground system for the power supply common mode to "push against".
Think of it as a little BPL system, with the telco or CATV lines one
conductor and the power lines another.
Connecting the Telco ground or antenna ground to NEW bypass capacitors on
the power line prevents the wonderful antenna and widely spaced
"counterpoise" from being at different potentials for RF.
The reason I stressed the safety ground is the .01 capacitors pass some
small current, and can "bite" the user if they are not grounded. This is
actually why manufacturers won't use such bypasses, it makes the case "hot"
with a very tiny current at 60 volts RMS. The safety ground connection
prevents that from happening, and prevents high current if a capacitor
shorts (extremely unlikely if you use the proper part, but still possible).
Its sole function is safety, otherwise it would not be needed.
> This could happen if the RF conducted on the power cord is
> common-mode. Both line and neutral can "drive" Safety Ground. And
> this is the reason Steve should be prepared to try ferrites too. They
> dissipate the RF instead of re-directing it.
Virtually all SMPS already have ferrites inside specifically for that
function. The reason they don't work well in every case is the required low
shunt impedance is missing. The manufacturer generally bypasses from power
line supply wire-to-wire, not from supply wire to chassis!
I'm not opposing ferrite cores, just reminding people that a choke doesn't
do much unless the impedance of the coke is very high compared to the shunt
impedance, and also a snap on bead has very little impedance with "one wire
pass". With little or no shunt capacitance, you don't have much of a
pi-filter!!! I have always found adding capacitors and grounding does far
more than adding chokes, mostly because most PS already have adequate chokes
inside...they simply lack bypass capacitors to the chassis.
By the way, the capacitors MUST be in a flame proof housing and you must use
this in a system with a safety ground. That's why I suggest buying a surge
protector that allows power line, CATV, and telephone lines to pass through.
I just get inside and add suitable .01 bypass caps on the power line and
telco lines. The CATV lines are already grounded to the neutral inside! I
replace telco line fuses with small molded 470uH chokes. An additional
benefit is all this work keeps RF out of the device, and greatly improves
lightning protection.
Few people will object to installing something that looks professional and
has improved lightning protection.
73 Tom
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|