A UHF+VSWRBridge

Being able to measure VSWR at UHF and microwave

fteguencies can aid those homebrew projects

By Paul Wade, N1BWT

ams frequently need to measure VSWR. This is paramount for antennas, a s i t is

the only parameter conveniently measured, but i t is also useful in rnany
homebrew projects.
Most of the VSWR meters we use for
antennas are based on directional couplers and require a significant amount
of power for a useful reading. Many
electronic devices are intolerant of
t h a t much power, and sometimes the
frequency involved is outside the ham
bands where i t is undesirable a s well
a s illegal to use transmitter power
levels.
A simple resistive VSWR bridge can
give good results over a wide frequency range with milliwatts of
power. I found a nice surplus unit for
the 2 to 12-GHz range which works so

161 Center Road

Shirley, MA 01464

well t h a t I wanted one for lower frequencies. I saw WlAIM using a bridge
to trim the phasing harness for his
EME array, and he reminded me that
Joe Reisert, W l J R , had described one
some years ago.' I later visited Joe a t
Antennaco and saw the origlnal unit
still being used to test antennas.
The W l J R unit is easily built in a
small aluminum box and works up to
about 450 MHz. He pointed out t h a t
the technique could be extended to
several GHz-just what I was looking
for. I decided to build one a s small a s
I could using inexpensive chip resistors and capacitors t h a t are now commonly available.
The circuit, shown schematically in
Fig 1, is quite simple. How it works is
more obvious if i t is redrawn a s a
Wheatstone bridge, Fig 2, with R 1
equal to R3.2 R2 is the reference load
'Notes appear on page 5

connected to 52, and R4 is the unknown impedance connected to 53; if

they are exactly the same, then the
voltage a t each end of R5 is the same,
so there is no detected output and the
VSWRis 1.0.This condition is referred
to a s having the bridge balanced. Any
difference between R3 and R4 unbalances the bridge and causes a n output
from the detector. Detected output is
proportional to bridge imbalance: the
higher the VSWR, the greater the output.
At high frequencies, all components
have inductance and capacitance.
Since we would like the bridge to opera t e a t a s high a frequency a s is
possible, i t is important to minimize
the L and C by making the bridge a s
small a s practical, and then to keep i t
balanced by making it symmetrical.
The smallest robust box t h a t I could
think of was n slice of X-band
waveguide. I sketched out the compo-

February 1995

nent dimensions to see if they would

fit-it was too tight for my fingers in
WR-90 waveguide. KlLPS had a spare
piece of WR-112, t h e next larger size
waveguide, with inner dimensions of
0.5 inches by 1.12 inches. The parts
a r e mounted on a small Teflon PC
board with the etching pattern shown
in Fin
" 3. Since all the lines a r e a s short
a s possible, dimensions a r e not critical; on the original board, I cut out the
pattern using a hobby knife. Note t h a t
layout and construction a r e symmetrical so t h a t 52 and 5 3 a r e electrically
identical and interchangeable.
Construction is straightforward.
The tinned PC board i s trimmed to fit
snugly in t h e waveguide, then t h e
SMA connectors, J 1 , 5 2 and 53, a r e
mounted and the center pins soldered
to the board to hold i t in place. Then
the unit is inverted to solder the
ground plane side to the inside of the
waveguide. After applying plenty of
flux, a ring of wire solder i s fitted
around the perimeter of the board,
ready to flow into place a s soon a s it
reaches t h e melting temperature. The
outside of t h e waveguide is heated
with a propane torch applied to t h e
side (right side in Fig 5 ) t h a t has no
SMA connector until t h e solder melts
and joins the PC board to t h e walls of
the waveguide. Don't be timid with t h e
torch-the idea is to get everything hot
quickly and remove the heat a s soon
a s the solder flows.
If waveguide i s not readily available, a suitable enclosure may be
fabricated from hobby brass, a s described by NJ2L.3 The pattern in Fig 3
extends past the waveguide dimensions to allow some flexibility in
packaging.
After the flux is scrubbed off, components a r e soldered in place. Component placement is shown in Fig 4 and
t h e photograph, Fig 5. A small wire
passes through t h e board to the BNC
connector and R8 underneath.
Operation of t h e bridge is also
straightforward. A modulated R F signal is applied to J 1 , and t h e detected
modulation, with amplitude proportional to VSWR, appears a t 54. Usually t h e modulation is a t 1 kHz, and
t h e detected output is connected to a
surplus SWR meter, such a s the H P
415. (The solid-state 415D and 415E
a r e more stable than earlier models,
but any of them work fine, a s do similar instruments from several other
manufacturers.) Since t h e SWR meter
is just a tuned audio amplifier with a
calibrated meter, any audio amplifier
driving a n output meter will do the job.
4

QEX

J7
Input

RF

J2

R5

J3

51
Unknown
Impedance

J4

17'7

Detected
Output

C3 ;
;
; 100 k

Fig 1-Schematic diagram of the VSWR bridge. All resistors and capacitors are
chip devices except for R8. Dl-Microwave mixer diode. R6, R7-10 to 15 kiL.

/A
Fig 2-The Wheatstone bridge is the
fundamental circuit within the
instrument.

Fig 4-Component

1.0 inch

Fig 3-Etching pattern for the printedcircuit board. Use Teflon board and trim
to fit the enclosure (%2 inch recommended thickness).

placement diagram for the VSWR bridge.

A good 50-R termination i s connected a t 5 2 a s t h e reference load: t h e

quality of t h e reference load i s import a n t since all other impedances a r e
compared to it. I n fact, we could
measure VSWR on a 75-R cable by
using a 75-R termination a t 52.
To calibrate a measurement, connect a coaxial short circuit a t 5 3 a n d
adjust t h e SWR meter scale for infinite
VSWR, or 100% power reflected. The
unknown impedance i s t h e n connected
a t 5 3 to measure i t s VSWR. However,
w h a t a typical SWR meter such a s t h e
HP415 displays is r e t u r n loss, t h e percentage of reflected power in dB. Most
microwave engineers u s e r e t u r n loss
directly, b u t if you a r e more comfortable with VSWR, t h i s i s t h e equation
for conversion (remember t h a t RL i s a
negative number):

Fig 5-The completed bridge assembly.

RI.

1+ 10%
SWR = mr,
~

Since we have tried to keep t h e

bridge balanced a n d symmetrical, a
good t e s t i s to p u t good loads on both
5 2 a n d 53, then swap them. The measured VSWR should be low (high
r e t u r n loss) a n d identical in both
cases.
Now t h a t we've gone to t h e effort of
using chip components a n d making
t h e VSWR bridge a s small a s possible,
how much have we gained? With good
50-R SMA terminations a t 5 2 a n d 53,
r e t u r n loss was g r e a t e r t h a n 3 0 d B
(VSWR c 1.2) from 10 MHz through
2304 MHz, while a t 3456 MHz, t h e
r e t u r n loss was 22 d B (VSWR = 1.9),so
performance i s degraded b u t still
usable. However, a t 5760 MHz i t was
worthless. So we have increased t h e

upper frequency limit to a t least five

times a s high a s t h e original version.
I t is possible t h a t t h i s style ofVSWR
bridge could be pushed even higher in
frequency by making i t even smaller,
using tiny ( a n d more expensive) microwave chip capacitors a n d resistors.

Wlll e blgger amp do the 1007

What about changlng the snfenne?

CAPMAN provldes answers In a user frlendly menu drlven

skywave analysis package with context-sensitive help & color graphics.
L ~ . ~10.
I I TM
nib-. usa m
S S M : 37 r l - :
rz
m m n = .IW NU. J m a WISE = 14a.o WY. NO. s m t = u . o
62.42 ~ c v r160.a9. m 1373.1 un 1316.3. n a ~ n ft a zr8~.:1o.la8
F
y
Local

...

-.-*--,

1U

1X

14

Ib

---

1U

:In

,
,
A
dU

ad

d4

..

u-.*
Y I l ~ . b t l I l ~
U r - - l w I ~ h hm I
~ m r r r t r cby -r.
ulnd bu ~ ~ t t o
r rr mrr.s
CUB t o
11, sm. 1a3 n - 1 1 a 1 1 1 t
< 2 > n - t r a d - ~ ~ rr .I*.I 0.1".
<'>
,#"'~,uuu,i.c~;,n~:,?;,~~::*
c n , N,
c u , 2.0 m x , cc, 3 . . m,. co, 7 . 2 m.. < k , 10.1 nu.,
<F, 14.a
10,
m a OO 2 1 . 2 m. ( 1 1 24.9 w. <
.
I
I>
l . 3 W' <I)
a * . . fm.
t - ~ r e~ ~
u t m u tm u F - ~ s h-r
s-1-cc

,<

-.

,I-

r:".,-.. <,

1m.1

:::: 5Z 2: ::::

%X :l :::!

1.37

. . ....

+ + +

' " .I

M:

8 R.II.bllltw
D-mllmt
111-

..
+

Entlre U S A 1-800-854-0547

00

15-17.

Know the best tlrnes a n d bands for skeds o r OX

1976, pp 50-56.

2Ryder. John D., Networks, Lines and

Fields. Prentice-Hall. 1955. pp 32-37.
3Healy, Rus. NJ2L, "Building Enclosures for
Microwave Circuits," QEX, June 1994, pp

Evaluate statlon changes before maklng them.

Ea.1

UWMUNMlKWS SKCIALISTS. M

'Reisert, Joe, W l JR, "Matching Techniques

for VHFlUHF Antennas," ham radio, July

O p t i m i z e your H F operation!

m: mtr

426 Wesl Tall Ave Orange. CA 9266542%

Local (714) 998-3321. FAX (714) 974.3420

Frenchtown, NJ 08825
Tel: 908-996-3584

Notes
Parts and boards are available from:
DownEast Microwave

SCP 1994

CCI Capacllot K~Icontatns365poeces.

bea ofeven,
10% value lrom lpI10 3 3 ~ 1CR-1Res~slorKII conlalns
154Opleces,lOea olevev 5% valuelrom lOn10 10megll
Sizes are 0805 and I206 Each k ~ ISt ONLY $4995 and
ava~lablelor lrnmedlals One Day Dellvev'
Order by toll-treephone. FAX, or mall We accept
VISA. MC. COD. or Pre.pald orders Company PO'S
accepted w~thapproved credll Call lor lree detatled
brochure

954 Rt 519

=:: :Wb,%

2:

:C?:,E
::

m.

:Mi,!%! 24

+ MUF, FOT, 3 4 ,RebbWlly, angles of l a k e 4 6

Ouick onestep forecest e n a ~ k .
enhrel. S u n b , mom using Es. E. Fl. F2-byes.
+ ~ b p t a and
y pmr outpvt gmph~csny.
+ Cuslome ALL parameters
+ IONCAP* used by wernrnenb and
broadcesten w & e
+ Use Fku/SSN 8 o p ~ n eK-Index.
l
.
+ Repulres 80386or above.
Use MlnlNeuElNecoutput for custom antennas.
+ Package $89,outsue USA tW.50
+ Caku!are 22 Skywave pammeters.
LUCAS RadiolKanaaroo Tabor Software. 2900 Valmonl Rd. Suite H. Boulder. CO 80301