On 9/3/20 2:52 PM, David Gilbert wrote:
I've done a BUNCH of Google searching using (among others) the search
phrases "dielectric properties of ferrites", "heating effects in
ferrites", "VSWR failures of baluns", etc and so far not found a clear
discussion of the topic. That first search phrase alone generated
1,230,000 hits and while I obviously didn't check them all I did check a
LOT of them.
Yes - you want to look for "permittivity" - that cuts down the hits a lot.
I remember, though, previous discussions here on TowerTalk where more
knowledgeable people than either you or I pointed out that damaging core
heating in a typical ferrite balun or common mode choke is far more
likely to be the result of dielectric losses due to high VSWR than it is
due to eddy current losses due to high current in the winding. I just
hoped that one of those people would know of a technical link that more
rigorously describes the condition.
What I do know is that ferrites are basically an iron-based ceramic with
other materials (usually other metals) included as dopants to get
various desired properties, such as mu or best frequency range. Being a
ceramic, ferrites are basically insulators with both a dielectric
constant and a loss tangent that results in current flow (and therefore
loss and heating) under RF voltage (E-field) excitation. That E-field
could logically be the result of a high VSWR ... just like the folks who
have stated same in the past.
As I remember, you were asking for others to do some studies that you
could have done yourself. I'm merely asking for a link if anybody knows
of one.
In a "choke" application (as opposed to a transformer application) -
let's think of a toroid.. where would the RF current "around the core"
(i.e. right angles to a central wire through the core) come from?
With multiple turns, I can see there being some asymmetry of the windings.
I've always thought that the heating was from magnetic loss, especially
if the choking impedance was too low (so the "choked current" isn't
reduced enough, so the equivalent I^2*R losses were high)
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|