EMRFD Message Archive 4896

Message Date From Subject
4896 2010-08-12 12:22:11 John Marshall Return Loss Bridges
On Aug 10, 2010 6:48 pm ((PDT)), "Phil Sittner" sittners@sbcglobal.net
wrote (in another thread):

> While clicking through the pages of this presentation I came across
> the RLB-18 design and assume it's a return loss bridge. Is anyone
> aware of this design and any of the pertinent details?
> regards,
> Phil kd6rm

The Return Loss Bridge Phil mentions is identical schematically to
Figures 7-41 and 70-42 in EMRFD (1st ed). These figures each show a
resistive bridge with a signal source feeding a reference and an
unknown impedance load through equal series resistors. A detector
measures the difference in voltage between the reference and unknown
ports of the bridge, and shows a null when the two impedances are equal.

I've built several bridges of this type or with a built-in diode
detector, like Figure 7-43. For these things to work well, the
reference and unknown arms must be as close to identical as you can
make them. This includes matching the fixed resistors in series with
the source, including their parasitic L's and C's, lead lengths, and
stray C's from the circuit layout. Electrical symmetry follows
physically symmetry, so using identical connectors for the reference
and unknown helps. It's also important that the detector connection be
well balanced with respect to ground, so that it doesn't load the
reference port any differently than the unknown port.

The bridges I built 30ish years ago all used leaded components, laid
out to give identically short lead lengths. This one from K8IQY is
similar although he uses a fixed resistor for the unknown:
< http://www.k8iqy.com/testequipment/returnlossbridge/returnlossbridge.htm

Today, using matched surface mount parts with microstripline
connections can take symmetry to a new level. This file from Sam
Wetterlin at:
< http://www.wetterlin.org/sam/Reflection/ManualReturnLoss.pdf >
shows precise SMD/stripline construction. The detector balun is quite
elaborate, using binocular cores placed inside toroid beads and
different ferrite mixes to cover a wide bandwidth. (Note the
similarity to the cascaded-sortabaluns-on-different-mix-toroids in the
original reference).

There are several other files on Sam's site at:
< http://www.wetterlin.org/sam/ > that show other bridges and how to
use them plus other interesting stuff.

John, KU4AF
Pittsboro, NC
4897 2010-08-12 17:03:38 Phil Sittner Re: Return Loss Bridges

Thank you for the info. Sam's notes are most interesting.

Phil kd6rm
----- Original Message -----
7316 2012-03-08 21:12:40 Kerry Return Loss Bridges
Wes' recent paper on RLBs is timely for me; I find bridges of all kinds very interesting.

I've built several RLBs over the years; as I learn more, the bridges improve.

My most recent effort, built a couple of years ago, works very well to 1.3 GHz (the limit of my reliable test equipment) but I've been thinking for months about a newer and better (I hope) design; Wes' paper has given me a few things to consider.

The balun/transformer is, of course, the key; other aspects are largely reliant on good construction & layout so it's the balun/transformer that I'm considering.

The purpose of this post is to explore Wes' comment to the effect that 43 material might not be ideal; I'm not sure why this might be so.

If co-axial line is used, as in the "string-of-beads" or "trombone" design, the differential currents are not affected by the core (or vice versa); the same ought to apply to carefully-made twisted line.

It's only common-mode currents that should concern us and it seems to me that a high resistive component of the total Z would not be a disadvantage.

Low core Q should reduce any tendency for resonance with various "strays" and would, on the face of it, be a good thing. 43 material should be good at the lower/middle end of a broad frequency range and 61 could "take over" at the high end.

Would Wes (or anyone) like to comment?