Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

How to design small GSM/DCS PCB antenna ?

Status
Not open for further replies.

yellowtooth

Full Member level 3
Joined
Aug 23, 2001
Messages
173
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,296
Activity points
1,626
pcb gsm antenna

I want to lay a GSM/DCS antenna directly to PCB.

So the size of it should not be too large.

Attachment is my design.

(red is top view, blue is bottom view)

The resonate frequency is just tunable, and can be the point I want.

But the bandwidth is too narrow.

What sould I do next ?

Any other desing ?

Thank U.


I have designed the IFA successfully for Bluetooth.

But the GSM&DCS frquency is too much lower than Bluetooth.


So, I think, the size maybe too large.

Is there any method to desing small IFA for GSM&DCS ?

Thank U.
 

gsm antenna design

This is a Meander Inverted-L type antenna. From my experience you will get very low antenna gain with this kind of antenna in GSM band (aprox. -8dBi, due to high Radiation Resistance).
GSM spec for antenna gain is much harder to meet compared to Bluetooth.
Inverted-F or PIFA are good choices for GSM antenna, but these designs require bigger space than yours.
 

gsm pcb antenna design

1. "This is a Meander Inverted-L type antenna. Inverted-F or PIFA are good choices for GSM antenna, but these designs require bigger space than yours."

How about a meande iverted F type antenna then ? Mabe the size can be small. But will the band width be broader ? Or the gain be larger ?

2. "you will get very low antenna gain with this kind of antenna in GSM band (aprox. -8dBi, due to high Radiation Resistance). "

Do U mean that both IFA(iverted F type antenna ) and ILA(Inverted-L type antenna) have poor gain ? Or IFA better than ILA ?


3. "GSM spec for antenna gain is much harder to meet compared to Bluetooth. "

The directivity for GSM antnenna should be better than Bluetooth. So though IFA is good for Bluetooth, it is bad for GSM. Am I right ?
 

pcb antenna design

I said that Meander Inverted-L antenna for GSM (900MHz) on PCB has poor gain.
Even a Meander Inverted-F will have low gain at these dimensions and frequency, but this one allow for a better matching compared with Inverted-L.

There are a few factors that affect radiated resistance in a meander type compared with other antenna types: loss in RF conductor resistance, eddy current loss, corona loss, ground resistance loss, etc.

GSM handset antenna shouldn’t be a directive antenna and its pattern shall be close to an omnidirectional pattern (in the limits to pass SAR).

Because GSM is a cellular network (Blootooth isn’t), cannot make too much compromise for antenna gain, because the required radiated power of the handset is very well specified in the specs (e.g. ETSI-TS 151 010).
 

gsm pcb antenna

I think, both Bluetooth and GSM should have omini-directional antenna.

So either gain can be high.

For the gain only can be high when it is directional.


And I want to know what should I do now.

Thank U.
 

pcb antenna

I don’t know who said that Bluetooth antenna should be directional, even if can be in a point-to-point application.
IFA is good for GSM900, but not Meander type. I arrived to this conclusion after a lot of time spent with this topology.
Probably the best option for GSM900 is PIFA, which anyway is chosen by the majority of phone manufacturers. PIFA has better BW and better SAR behavior compared with standard IFA.
 

gsm antenna pcb

I am so sorry now.

PIFA is too big for me.

IFA with no meander is to long for me. (L is about λ/4, 83mm for 900MHz, 41mm for 1800MHz)

So I have to use the meander, which has narrow band width.

God help me.

Added after 17 minutes:

I have a idea:

compose two meander together.

Maybe the band width can double.

What I do is just to tune the two resonant frequency a little different.

Am I right?
 

gsm antenna on pcb

To make the two antennas to resonate independently you need some space between them. Even if you get some second resonance, the gain still will be very low.
A good antenna shall be straight as possible in a line. When start to fold, bend, zigzag, or curve (especially in the high current areas) the gain goes dramatically down. The high currents areas shall keep away from other conductors (or second meander in your case).
 

small gsm antenna

One more question:

How to pronounce the word "PIFA" ?

[pi:fε] or [paifε] ?

Thank U

Added after 15 minutes:

vfone said:
To make the two antennas to resonate independently you need some space between them. Even if you get some second resonance, the gain still will be very low.
A good antenna shall be straight as possible in a line. When start to fold, bend, zigzag, or curve (especially in the high current areas) the gain goes dramatically down. The high currents areas shall keep away from other conductors (or second meander in your case).

Gain=Directivity*efficency

U said the Gain will be low, then what the major reason for that , the Directivity or the efficency ?

Added after 56 minutes:

Now, I want to feed the tow meander inverted L antennas only though one point.

So the distance between them can not be too far.

Another question:

If I feed the tow antennas just as dipole, i.e. coaxial + connect to one, coaxial - connect to another, and the - on the other side of the coaxial connect to GND.

Does the GND below the antennas still have big impat on the antennas ?

Will the badn width be broarder ?
 

pcb gsm antennas

The gain is low because the efficiency is low.
If you change the topology to a dipole there will not be anymore a monopole (inverted L or F are monopoles). Monopoles needs a ground in the nearby to create the mirror image of the element, when the dipole don’t need and don’t like a ground in vicinity. This is just antenna basics.
 

make gsm antenna

I test the Ant in a TEM cell (produced by TESTCOMM in Korea), whose input and output are just from and to a network analyser .

The S11(VSWR) is just good, but the S21 is too bad.

So I will give up, and use the IFA without meander, which I tink has better gain performance.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top