WLAN 802.11b/g Bi-quad antenna: building and testing

Report on reproducing Trevor Marshall's BiQuad Antenna.

Antenna specifications: 11dBi, wide band

For original article see: http://trevormarshall.com/biquad.htm


Antenna wire: lenght 244 mm, diameter 1.2 mm, aluminium
Reflector: circuitboard 123 x 123 mm
Cable: N-female 300mm test-cable MIL-STD RG-58C (semi-rigid)




Complete antenna.

Notes on building:
The center frequency is very sensitive to how you solder the ends of the quad loops to the coax. After multiple attempts it was determined that even a half a millimeter difference will move the center frequency by as much as 50 MHz. So be prepared for a lot of "measure, resolder, measure, resolder ..."

The feedpoint:
one branch of quad to coax center conductor, and the other to the shield (ground).

The feedpoint is the most ‘ugly’ place in the antenna. Try to keep the gap as short as practicable and the crossing lines together in symmetry - however be prepared to resolder a couple of times to find the best configuration. Note that, to allow for a gap between the feedpoint, the ‘quad-ring’ will be distorted somewhat - however try to keep the quad overall lenght the same 244mm.

When you measure the returnloss/SWR and find that the center frequency is off where you want it, you have to make the quad-ring 1mm shorter or longer by adjusting how you solder the quad ends to the coax. If you find that the returnloss/SWR figure isn't 'sharp' or is distorted then you probably have some non-symmetry at the feedpoint.

Spot on:
Note that the SWR at center frequency isn't actually as important as on the channels 1 and 11. You want to center to be where it dips those frequencies the most.


Test report:

Equipment:

Test conditions:
The analyser's tracking generator and separate return loss brigde were used to test the return loss from the test port without the antenna attached. This level was set as zero reference for the measurement markers. The antenna was suspended with cardboard 1 meter above ground with a minimum of 5 meters to adjacent walls and the ceiling. The antenna was moved around to average out reflections and other anomalies caused by the less than perfect measuring conditions.

Results:
Marker 1: 2412 MHz (Channel  1), Return loss -23dB = SWR 1.2
Marker 2: 2437 MHz (Channel  6), Return loss -33dB = SWR 1.1
Marker 3: 2462 MHz (Channel 11), Return loss -23dB = SWR 1.2

Notes on measuring:
No, you don't need a billion dollar network analyser to do antenna measurements. Very good results can be achieved with a ordinary spectrum analyser with tracking generator option. All you need is a good quality return loss bridge and some test cables and adapters. At 2.4GHz your equipment needs to be rated for atleast 3GHz operation and the cables and adapters must have low loss and good SWR at these frequencies. Using poor cables or adapters in reflection measurements will give you 'better' results than reality would warrant. Remember that in reflection loss measurements you are measuring the reflected energy from the antenna, which at good SWRs will be very small fraction of the tracking generators output.

Here's a small table for converting reflection loss and SWR: For example if the reflection loss is -20dB this means 1% of the energy transmitter is reflected back. This is equivalent to an SWR of 1,222 which is very good.

Ref dB SWR Ref Coff. ^ -1   Ref dB SWR Ref Coff. ^ -1
-1 17,391 0,7943 1,26   -21 1,196 0,0079 125,89
-2 8,724 0,6310 1,58   -22 1,173 0,0063 158,49
-3 5,848 0,5012 2,00   -23 1,152 0,0050 199,53
-4 4,419 0,3981 2,51   -24 1,135 0,0040 251,19
-5 3,570 0,3162 3,16   -25 1,119 0,0032 316,23
-6 3,010 0,2512 3,98   -26 1,106 0,0025 398,11
-7 2,615 0,1995 5,01   -27 1,094 0,0020 501,19
-8 2,323 0,1585 6,31   -28 1,083 0,0016 630,96
-9 2,100 0,1259 7,94   -29 1,074 0,0013 794,33
-10 1,925 0,1000 10,00   -30 1,065 0,0010 1000,00
-11 1,785 0,0794 12,59   -31 1,058 0,0008 1258,93
-12 1,671 0,0631 15,85   -32 1,052 0,0006 1584,89
-13 1,577 0,0501 19,95   -33 1,046 0,0005 1995,26
-14 1,499 0,0398 25,12   -34 1,041 0,0004 2511,89
-15 1,433 0,0316 31,62   -35 1,036 0,0003 3162,28
-16 1,377 0,0251 39,81   -36 1,032 0,0003 3981,07
-17 1,329 0,0200 50,12   -37 1,029 0,0002 5011,87
-18 1,288 0,0158 63,10   -38 1,025 0,0002 6309,57
-19 1,253 0,0126 79,43   -39 1,023 0,0001 7943,28
-20 1,222 0,0100 100,00   -40 1,020 0,0001 10000,00

( VSWR = 1-SQRT(Reflected/Forward) : 1+(SQRT(Reflected/Forward) )


Contact:

Miikka Raninen (OH3GPJ), Tampere, Finland

Created on ... June 29, 2006