[Boatanchors] RF Ammeter Design and Construction

[J. Forster]

Background: A question was posted about an RF Ammeter and I suggested some
DC tests, which produced odd results...   if you applied a DC current in
one direction the meter read one value; if the current was reversed, a
dramatically different value.

In thinking about this, the meter design must have been different from
conventional.

A conventional design looks sort of like this:

O--------o-------O    Main RF path
        / \
        | |    Thermocouple leads
       meter

The ones I've seen have the two Thermocouple leads welded together and
then attached to the main resistance wire with a small drop of cement.
This design should be completely independant of the current flow in the
main path.

Another design?

Based on the measurements, about the only plausible design fitting the
data is one like this:

O------o---o-----O    Main RF path
      /     \
     |       |    Thermocouple leads
      meter

It may look funky, but should work, and explain the measurements. From the
meter circuit's point of view, the two series couples (left TC to main &
main to TC right) appears as just a single couple, if the main line is
uniform. The middle section of dissimilar metals cancels out.

However, if there is a DC component in the main line, ther IR drop between
the couple points will add or subtract from the thermoelectrically
generated voltage, giving different readings depending on the direction of
the DC current.

This is strange to me, and I've never seen it discussed anywhere.

Does anybody know of references. It is important because it is hard to
calibrate RF Ammeters at RF, and this rules out the use of a thermocouple
as a transfer standard between AC and DC.

Best,

-John

======================