On Fri, 28 Jan 2005 16:55:33 +1300, MikeN
wrote:

6. Replace the ferrite bead with a resistor, and apply DC voltage
same as E to get the same temperature rise in the same time period.

Per your earlier use of an RF detector, you have to first consider the
conversion factor (was the meter peak reading or average?). Why
bother, you already have a power source (the HT) you already have a
detector (your same simple detector). No conversion necessary.

Change the resistance value as necessary to get the same temperature
rise over same time.

7. Dummy load is now dissipating the same power as the ferrite
core did.

8. Calculate the impedance from Z=E^2 / W.

Hi Mike,

Step 8. is unnecessary given step 7. You only want to know the
ferrite R which is directly obtainable from the Resistor setting (or
Resistor choice). You don't even need to compute power anymore as
that has fallen out of the equation in the method you describe -
which, by the way, is a good example of crafting a solution. It shows
you simplifying my stark description of caloric measurement to instead
engage in bolometrics (comparison of heating). This method is
probably superior in simplicity and results would easily be within 20%
(which is not shabby for UHF).

Your introduction of a resistor is also an example of what
Metrologists call a "transfer standard." It is an example of using
the "substitution method." Your only concern is that the resistor
present a true resistance and not some complex Z. In other words it
should match the feed, and not offer much stray X. With that said, it
becomes a tougher problem (but then you needed to do the same thing
with the actual Z of the ferrite) that is, matching. It is arguable
that they would both mismatch equally (and given the Power term is
canceled, match is not particularly necessary). Everything here
depends on your re-obtaining identical indications.

This discussion reveals the cost of absolute determinations.

To increase the success it behooves you to up the power to cut down on
environmental temperature biasing the experiment (also cuts down on
other subtle influences like the difference in mass of heated
samples).

There are also indirect methods which can tolerate far more
imprecision. Ferrites are composed with bulk properties that have
frequency dependencies. These properties, however, vary quite
smoothly and slowly across great ranges of frequency. They also
exhibit distinctive family properties. The different grades of
Ferrites react with peak Resistances in different bands, but for our
purposes one family of Ferrites can be quite useful across a
significant percentage of bandwidth. Consult:
http://bytemark.com/products/ferrmat.htm
Unfortunately, Bytemark.com has fallen short of complete
documentation. They offer a link to illustrate Z over F, but it is a
dead link (and has been for years).

However, by this one page alone you can discern the family
characteristics I speak of. Your best hope is that your beads are
composed of something like type 43 or 64 instead of type 61 or 73.

An indirect method would be to measure the bead in series with a good
resistor - at two or three frequencies. HF, VHF and UHF would be
eminently suitable. If the bead shows a higher R at UHF, this trend
would tend to support an assumption you are have a suitable material
type. Both types 43 and 64 should exhibit useful resistances in both
bands. The slope frequency characteristics of these materials easily
span both higher bands.

Let's put some useful context to this. For any bead that snuggly fits
over the jacket of an RG-58 cable the following Rs should be seen for:

Material HF VHF UHF
75 ~20 ~10 ~5
73 | 77 25-30 ~20 ~15
43 ~20 ~32 ~30
64 ~5 ~30 ~35

This should put material identification within reach.

73's
Richard Clark, KB7QHC