[time-nuts] Noise and non-linear behaviour of ferrite transformers
kb8tq at n1k.org
Sun Jul 20 22:20:30 EDT 2014
It is impractical to make coax that has a shield thickness of 1/3”. Even if you do, it’s not going to be very flexible. For a real world system that needs good isolation, coax is not the way to go below 100 KHz. There are a few other issues that come up, but skin depth is a big part of the problem.
Another part of the equation is (as Bert points out in another thread) “how good do you need?”. Skin depth simply the point that you have knocked out 2/3 of the current. That probably isn’t what you are after when you ask for “good isolation”. The “inside” of the coax should be below 170 dbm/ Hz to be “quiet” when terminated. If you have -70 dbm / Hz noise signals running around here and there, you need quite a bit of isolation. You might have a spur spec rather than a noise floor spec to meet and that would give you different numbers to go after. In most cases you will need multiple skin depths (like 10 or more) to get the job done in a noisy environment.
On Jul 20, 2014, at 9:10 PM, Chuck Harris <cfharris at erols.com> wrote:
> I'm not sure what you are saying.
> skin depth = (2.6/sqrt(fhz))inches for copper.
> So, at 60Hz, skin depth = 0.336 inches.
> and at 100KHz, skin depth = 0.008 inches.
> and at 1MHz, skin depth = 0.0026 inches.
> Are you saying that at 60Hz, because the
> skin depth is deeper than the coax shield is
> thick, that current passing through the outside
> of the shield will induce voltage inside of
> the shield, and that at say, 100KHz where the
> skin depth is a little less than the shield
> thickness, or at 1 MHz, where the skin depth
> is only a small fraction of the thickness of
> the shield that it won't?
> Or something else?
> -Chuck Harris
> Bob Camp wrote:
>> The “coax is an antenna” problem comes in well before you get to DC. Even with no
>> transformer involved, the skin depth of the coax shield gives up well above 60 Hz
>> (and likely well above 100 KHz). If you want to do full isolation over a very wide
>> range you need some combination of shielding and balanced lines.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts