434MHZ PCB antenna design help

Hi

I'm using following RF modules in my project and need urgent help on designing antennas. My entire project is halt because of the antenna problem. I cant send transmission to more than 1.5m without corruption. I'm currently using 30cm long pieces of wires hanging in open air outside the enclosure.
So I'm looking for a PCB antenna.

Receiver
**broken link removed**

Transmitter
**broken link removed**

I have tight space consideration and need to design a PCB based antenna.
Can some one suggest me a proper PCB antenna type and a design procedure? I.E. how to calculate length, width angles etc of the antenna.

It would be much better if antenna does not require any tuning items a like capacitors or inductors.

and antenna will be on a separate copper board because rf modules are pre made without antennas

I heard meander antennas and inverted F antennas are good for this but I couldn't find any document on how to design a such antenna for my 434MHz system.

Thank you very much

It would be much better if antenna does not require any tuning items a like capacitors or inductors.

Click to expand...

Electrical small antennas have a radiation resistance considerably below the 50 ohm impedance of most RF modules and won't hardly achieve good performance without a matching network.

Besides antenna textbooks like Ballanis, Antenna Theory, you'll find a lot of information in manufacturer application notes, e.g. TI AN003 SRD Antennas **broken link removed**
or other literature under Analog, Embedded Processing, Semiconductor Company, Texas Instruments -> Wireless Connectivity -> Technical Documents

hi FvM

The book and pdf does not contain design procedures for meander or F antennas.
I'm quite low on time and looking for quick equations where I can put values and obtain antenna design values. I dont have time to read entire book,coz this is for my final year project and, I have to finish this part quickly because lot work is still has to be done. If you can please provide me with equations or a design tutorial please.
thank you.

As far as I remember, AN003 is giving the geometry for the compared antennas. There are newer TI application notes with more design examples, e.g. DN031.

Hello,

30 cm wire length is a bad choice as this has high input impedance so you have huge mismatch towards 50 Ohms. A descent antenna will increase the useful range.

What range do you need? When you have 1.5m now, even with a good omni-directional antenna you will not get 10m.

Try to find just the antenna type that my fit your needs and scale it to your frequency of interest to get a first guess.

The scaling will not give you the exact antenna size (as PCB thickness cannot always be scaled and vicinity of other objects will affect the antenna tuning and impedance).

As FvM said, small antennas aren't 50 Ohms devices. In addition, small antennas have narrow bandwidth, so tuning (by changing the geometry) is very important. Narrow band means also high circulating current and relative high voltage, hence large field strength. Therefore also you have to consider the influence of nearby object as these may result in decreased antenna efficiency due to dissipation. Making your antenna too small will result in bad radiation efficiency due to dissipation in nearby objects (think of PCB material, nylon spacers, etc) and copper loss.

Common mode issues are frequently forgotten to take into account, but can make the difference.

To be honest, if you need very small size (thinks of several cm), you don't have the knowledge, and you need it fast, hire an expert and have access to a (single port) VNA.

It is not too likely that it is your antenna. These types of receivers typically have a pretty good dynamic range, and even with a mismatched antenna you should get more than 1.5 meters.

You could use a wire or pcb trace with a series inductor tuning it at its base. You choose a wire/trace length, and then vary the inductor value until it is tuned.

These OOK receivers sometimes have trouble with their AGC settings. You might have TOO much power recevied. Will it work 5 meters away, but not work at 1.5 m? Sometimes you have to vary the components controlling the AGC attack time to match your data rate. The AGC might be too fast, and cutting out some of the data.

You could have other problems, like bad center frequency.

biff44: I agree with you, there must be something else. If the products conform to the datasheet (that is sensitivity <-95 dBm and +9dBm transmitting power), even a wire of about 8 cm (on both units and some other wire or groundplane as counterpoise) would result in >> 1.5m range. Until now I didn't check the datasheets.

If you have access to another signal source (that you can OOK), you may test the receiver in a controlled manner to see if something is wrong.

Do other 433 SRD units (from other brand) have the expected range (to rule out very strong interference)?

FvM said:

As far as I remember, AN003 is giving the geometry for the compared antennas. There are newer TI application notes with more design examples, e.g. DN031.

Click to expand...

I shall check them

biff44 said:

It is not too likely that it is your antenna. These types of receivers typically have a pretty good dynamic range, and even with a mismatched antenna you should get more than 1.5 meters.

You could use a wire or pcb trace with a series inductor tuning it at its base. You choose a wire/trace length, and then vary the inductor value until it is tuned.

These OOK receivers sometimes have trouble with their AGC settings. You might have TOO much power recevied. Will it work 5 meters away, but not work at 1.5 m? Sometimes you have to vary the components controlling the AGC attack time to match your data rate. The AGC might be too fast, and cutting out some of the data.

You could have other problems, like bad center frequency.

Click to expand...

when distance is more than 1.5m data get corrupts and at about 4m data reception is lost

WimRFP said:

Hello,

30 cm wire length is a bad choice as this has high input impedance so you have huge mismatch towards 50 Ohms. A descent antenna will increase the useful range.

What range do you need? When you have 1.5m now, even with a good omni-directional antenna you will not get 10m.

Try to find just the antenna type that my fit your needs and scale it to your frequency of interest to get a first guess.

The scaling will not give you the exact antenna size (as PCB thickness cannot always be scaled and vicinity of other objects will affect the antenna tuning and impedance).

As FvM said, small antennas aren't 50 Ohms devices. In addition, small antennas have narrow bandwidth, so tuning (by changing the geometry) is very important. Narrow band means also high circulating current and relative high voltage, hence large field strength. Therefore also you have to consider the influence of nearby object as these may result in decreased antenna efficiency due to dissipation. Making your antenna too small will result in bad radiation efficiency due to dissipation in nearby objects (think of PCB material, nylon spacers, etc) and copper loss.

Common mode issues are frequently forgotten to take into account, but can make the difference.

To be honest, if you need very small size (thinks of several cm), you don't have the knowledge, and you need it fast, hire an expert and have access to a (single port) VNA.

Click to expand...

hi,

is there any way to measure the antenna impedance without using a network analyzer?

say I want to measure the impedance of my 30cm wire, how can I measure it?

and as you say can I use (lambda/4) long wire with series inductor to get 50 ohm impedance? or use a copper path + series inductor to achieve this?
if So how to calculate copper path length and required inductor value?

WimRFP

Re: 434MHZ PCB antenna design help
biff44: I agree with you, there must be something else. If the products conform to the datasheet (that is sensitivity <-95 dBm and +9dBm transmitting power), even a wire of about 8 cm (on both units and some other wire or ground plane as counterpoise) would result in >> 1.5m range. Until now I didn't check the datasheets.

If you have access to another signal source (that you can OOK), you may test the receiver in a controlled manner to see if something is wrong.

Do other 433 SRD units (from other brand) have the expected range (to rule out very strong interference)?

Click to expand...

well there is no ground plane. RF module doesn't have a ground plane. what I dont like about monopole is that I have to use a 17cm antenna + large ground plane, which will become quite annoying and bulky.
I dont have any other brand modules

What is the actual space available? Can you provide us a drawing or picture with everything in it (RF module, main PCB, battery, housing and wiring) and the available space for the antenna?

Regarding measurement of impedance, yes you can with a slotted line or a home made microstrip. By first probing the standing wave pattern along the slotted line/microstrip you know the cable loss and the position of the reference plane. With the DUT connected, you can determine the actual reflection coefficient based on the standing wave pattern. It is elaborate, but it can give good result. It is one of my favourites during courses.

If you are only interested in VSWR, you could use a directional coupler, you may check Ebay or other trade site for a coupler for your frequency range. I sometimes use a 2…100 MHz coupler from mini circuits. If you don't have a spectrum analyzer or 400 MHz capable oscilloscope, you could use a diode probe (power probe). Other option is a downconverter with LO and viewing the IF frequency on an oscilloscope. another option is to build a wheatstone bridge with floating diode detector as shown here: Homebrew SWR meter for VHF, UHF and SHF (including Wifi ISM Band). There are many varieties. When using SMD and good RF construction, performance can be good well over 2 GHz.

As far as I know, most HAM (licensed radio amateur) VNA's do not include 433 MHz, but you may give it a try

A complete other issue (even when having the best VNA) is that when you connect the feed cable to your circuit, the shield of the cable becomes more or less part of the antenna. Especially with circuit with small ground planes, this will severally affect the measurement when using monopole antennas. The impedance is not the value for the monopole alone, but a combined (not summed) value for both the (small) ground plane and the monopole.

To avoid the cable coupling issue, you need common mode decoupling. This can be done with ferrites (yes also at 433 MHz) and/or resonating/quarter wave traps or other means. With very small ground planes and quarter wave whips, it is hardly possible to get sufficient decoupling. I then us the whip itself as ground for the feed cable.

Regarding the design itself, if you don't have any ground/counterpoise provision, you may consider a tuned loop antenna. In case of a hand held device (for example a remote control), you may use the hand as capacitive groundplane.

Without suitable measurement tools, reproducing a 433 MHz PCB antenna design example, e.g. from a TI application note is probably teh best way. In addition you can use a dipole with diode detector as fieldstrength monitor, that should work in the transmitter's vicinity.

Another possible reason for transmission failure may be in not using suitable data encoding and decoding. I understand, that your modules have direct base band signal in- respectively outputs and don't implement any data encoding. For a reliable data transmission, you would need e.g. manchester encoding, sync preambles, and a kind of integrity check, e.g. CRC.

If no suitable measurement tools, with hfss simulation is a good choice