I have built a design out the ARRL handbook using a toroid transformer,
and SM 100 ohm resistors in parallel to make up the 50 ohm sections. I
could get 40 dB return loss to about 50 MHz, when matching a 50 ohm
load. This gives an idea of the accuracy of your bridge, ie the higher
the return loss in this situation, the better the bridge, assuming an
accurate 50 ohm load. I used a spectrum analyser as the detector and the
venerable HP8640B for the source.
You can even measure small inductances (I got reasonable accuracy to
about 100nH) using a series RLC combination on the end of a length of
coax as the unknown load. (ie do it remotely!) R=50, C=is known, L is
unknown
At some frequency there is a max return loss. At this freq XL=XC and
X(LC)~0, assuming perfect components. Thus the load is 50 ohms. You can
work out L knowing C

Richard

Henry Kolesnik wrote:
Does anyone recall seeing an article for constructing a return loss bridge
in RF Design in late 80s or early 90s? I'd like to see if I can get a
scan, I've been bitten by the RLB bug.
tnx
Hank WD5JFR

Richard Hosking wrote:
I have built a design out the ARRL handbook using a toroid transformer,
and SM 100 ohm resistors in parallel to make up the 50 ohm sections. I
could get 40 dB return loss to about 50 MHz, when matching a 50 ohm
load. This gives an idea of the accuracy of your bridge, ie the higher
the return loss in this situation, the better the bridge, assuming an
accurate 50 ohm load. I used a spectrum analyser as the detector

That also answers most of Reg's question about the difference between an
RLB and an SWR bridge.

As Reg has often pointed out, the so-called "SWR bridge" is actually a
reflection coefficient meter and it generally includes its own RF
voltmeter.

An RLB generally doesn't, but has an RF output port for the reflected
signal. That port is intended to be connected to a sensitive external RF
power detector - such as Richard's spectrum analyser - which can
accurately measure reflected signals down to -40dB... or even lower if
you consider such readings meaningful.

Another difference from the normal SWR bridge is that an RLB is always a
true bridge, with separate 'unknown' and 'reference' ports.

Normally your standard Zo load lives on the reference port, but if you
remove it, you can then use the RLB to compare *any* two impedances that
are intended to be the same (for example the two halves of a stacked
yagi array). You can see how alike the two impedances are, without
actually having to measure the value of either one... a very handy
feature.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)

http://www.ifwtech.co.uk/g3sek