"Arfa Daily" wrote in message

"Rich Wilson" wrote in message
...

"Glenn Richards" wrote
in message ...
Rich Wilson wrote:

If you're going to upgrade your power cable you really
ought to do it all the way back to the substation,
otherwise it's a bit pointless...

Well that was my initial thought... but apparently not.

In a conversation I had with a friend, we determined
that power cables are good at carrying low frequencies,
and poor at carrying higher frequencies. So RFI picked
up at the substation won't make it to your house, but
interference sources in your house will cause RFI to
reach your kit.

What I want to know is how exactly the RFI gets through
all the smoothing capacitors and so on in the rectifier.
I've got a little headphone amp here with enough
capacitance to not notice, say, a 2-second cutout in its
power supply, so I fail to see how any audible frequency
could get through.
Electrolytic capacitors generally have very poor
characteristics at high frequencies.

Unh, they used to be this way, maybe 20-30 years ago.

This is why you will
often find them bypassed with a small value polyester or
other film capacitor.

See former comments about 20-30 years ago. These days the capacitors may
still go in, but they are there for marketing, not technical purposes.

If you think about it, its trivial to eliminate inductance, other than that
due to lead length, from an electrolytic.

A large electrolytic across a
supply as a smoothing cap will, as you say, happily
supply the amplifier's needs for a couple of seconds, but
as far as RF goes, it may as well not be there.

Ignores the fact that the major trap for RF in a power supply is the power
transformer.

Also, you're confusing RF with audio. A signal is not really
considered to be RF until it reaches 60 or 70 kHz or so.
RF is of course, not directly audible anyway, even if it
gets into your amp, but if it had enough level, it could
conceivably drive early stages into non linearity, in
which case, its effect may become audible.

The most likely way that RF enters power amps is thorugh the input
terminals. Most of the time there is a simple small series resistor and
parallel cap at the input that deals with this issue.

Also, power amps should be able to defend themselves. The amp itself should
have a low-pass response characteristic that extends up to many megahertz.
Above that, the simple input filter should suffice.

In contrast, although you might consider impulse
interference from say a drill motor, to be an audible
frequency, the very fast rise times to high peak
amplitudes, will again bypass any electrolytics, as
though they weren't there, and may then become audible as
' interference '

This discussion about large electrolytics with excess inductance is old,
obsolete news.