monopole antenna matching

[adnan012]

I tried to design monopole antenna for 900MHz using HFSS.
My antenna dimensions are

Radiator length 90mm
Radiator radius 0.5mm
Dielectric radius 1.2mm

Simulation gives me antenna impedance of 79ohm with some inductive part.
Do i need to match the antenna impedance to 50 Ohm.

When i connect the antenna to VECTOR NETWORK analyzer (VNA), its vswr was not too bad. But there was confusion that VNA shows slightly different impedance from that i got from HFSS.

I have a 1 watt GSM MODEM module with a small antenna. When i connected this antenna to VNA , it shows large vale of VSWR. I also tried GSM antenna from a 1 WATT GSM jammer the result was same. Why such type of antennas shows different impedance from 50 ohm , and they still work.

I tested this antenna. The amplifier in GSM modem also have 50 Ohm i/o impdance.

**broken link removed**

 

[WimRFP]

Hello,

I would make it somewhat shorter, to remove the inductive part. Re(Z) will also go down a bit. If you want some better wide band performance, increase the wire radius.

Note that good operation of the monopole strongly depends on de groundplane you use.

Many modems are very tolerant to bad SWR as in real world many such antennas have SWR > 2. But try to keep SWR <2 (that means |s11| < 0.33333).

 

[adnan012]

thanks for reply.

i am using inner conductor of RG142 COAX CABLE.

I take an arbitrary length of 1 mm dia from rg142 coax wire . soldered it on a n-type connector and mounted it on a 200 by 200 mm fr4 ground plane . instead of cutting antenna wire i make spiral of multiple turns at the soldered edge . by varying number of turns and then by compressing and expanding it s turns, i brought VSWR under 1.9 from 880MHz to 915MHz band.

now the lenghts are

170mm straight length of radiator.
30mm spiral length with 9 turns with 4 mm external diameter.

may i expect that it will work?

 

[WimRFP]

Hello Adnan,

Your setup is not clear to me, 9 turns of 4mm diameter would result in about 110 mm.

For me it is strange that your VSWR is under 1.9, as (when I understand you well), the spiral has a length of 30mm (measured from the groundplane?).

A total radiator length of 30 mm will have significant change. It reduces the radiation resistance significantly. Don't look strange when Re(Z) drops to some ohms. You will also have less useful bandwidth.

Maybe I didn't understand what you meant.

 

[adnan012]

i measured carefully

8 to 9 turns of spiral /coil.
spiral / coil length 31mm.
spiral / coil external diameter is 4mm.
the gap between to adjacent turns is about 2 mm but it is not uniform.
170 mm of radiator length.

 

[WimRFP]

Hello,

The picture makes everything clear. You can forget my previous posting, as the assumption are wrong.

Your radiating part (170mm) is about 0.5 lambda long. When you would feed that directly, the VSWR would be very bad because of the high input impedance (you are in a voltage maximum).

The coil acts as an electrically quarter wave section. This converts the high input impedance of the 170mm radiator into a reasonable value (seen from 50 Ohms). In fact you are in a current maximum now.

The radiation from the spring coil/spiral will be small as the physical length is small with respect to the radiator length and the current is not much larger then in the 170mm radiator.

This antenna should have somewhat more gain (horizontal plane) with respect to the quarter wave antenna (over the same groundplane).

Why would you use this more complex arrangement?

 

[adnan012]

Thanks for your reply.

I am working on a high power ( 60 watt ) gsm amplifier. i cant find any good antenna from local market. This is why i am trying to construct my own antenna. i found that 5/8 antenna has close impedance to 50 ohm. is there any good solution for my problem? i also want to construct antennas in whole vhf and uhf range.

last day i derived this antenna by an amplifier (which i take from an old mobile phone). i placed the receiving antenna one wavelength away away and connected it to vna receiving port. i measured 10dbm power at 800 mhz. (i modified the antenna slightly for 800mhz).

 

[WimRFP]

Hello,

A 5/8 lambda radiator over large groundplane has Re(Z) around 50 Ohms, but Im(Z) < 0 (capacitive part), so you need an inductor (or piece of transmission line) to get it 50 + j0 Ohms.

If it is just to test something, why not making a thick quarter wave radiator over a ground plane? The thicker the radiator, the lower the fieldstrength at the metal-air interface and the wider the bandwidth. 60 W PEP should be no problem for a simple quarter wave whip over groundplane. Probably Re(Z) will be below 50 Ohms. Just extend it until Re(Z) = 50 Ohms and use a series capacitor to counteract the inductive part. You may also use several sloping quarter wave radials to raise Re(Z).

Regarding gain, The quoted additional gain for a 5/8wave over a half wave antenna is only valid for an infinite ground plane.

 

[adnan012]

Sorry for late reply.

I have tried thick quarter wave and 5/8 monopole. Both are working for me at 15 watt output power. Soon I will share complete measurements.

 

[adnan012]

I was busy in prototyping and other tasks.

i tested my amplifier and antenna using the setup given in the attached file.

vector network analyzer feed driver stage with 0dbm power. driver stage and power stage boost the power to 10-15 watts. the transmitting antenna is 5/8 lambda with spiral (inductor) match. receiving antenna is placed 12 inch or 0.3 meter away from the TX antenna. Rx antenna is lambda/4 . both antenna have 200*200 mm fr4 ground plane. on the receiving port of network analyzer i received 15 to 20 dbm of power from 880 to 915MHz band. Then i use the free space loss equation given at

LearningMeasure.com - Friis Transmission Equation Calculator

It tells me that transmitted power was 10 to 15 watts.

I want to know about the polarization of my antennas , and how monopole ntennas are polarized. and what is its significance.

 

[FvM]

Your setup is clearly near field and not suitable to determine antenna performance, I fear. You should go for minimum 3 m distance and preferably use a calibrated measurement antenna, e.g. a dipole for the receiving side. A lambda/4 with ground plane is introducing a too complex radiation characteristic.

By basic EM relations, linear antennas can't have but a linear polarization. If the ground plane is rotation symmetrical, it won't change this.

 

[adnan012]

How can i polarize antenna vertically or horizontally. Does it depend on feeding point type?

 

[FvM]

I depends on the field form of the antenna. A vertical monopole or dipole has always pure vertical polarization. Different polarization options exist for 2-D antennas, e.g. patches.

 

[adnan012]

Thanks a lot. I will increase the distance between the antennas to see the effect.

Here is a gsm antenna with variable polarization. Is it monopole type or something else?
Dual Polarization Panel GSM Antenna (ARP-890-960-21) - China Antenna,Wimax Antenna,Wifi Antenna in Antenna

 

[FvM]

Technical GSM antennas are usually built as arrays to achieve the intended directional characteristsic. For dual polarization, you have something like cross dipoles or special patch antennas.

 

[adnan012]

Does the cell phone antenna has polarization and directivity.