RFM22B antenna design problem

[hamed8419215]

hi.
i am using RFM22B-433-D (an RF module base on SI4432-revB1 that operate in 433MHz band and has DIP package).
DIP package of RFM22B has a little ground plane (with respect to SMD packages). and under some of the components, there is not any ground plane.
and also my circuit PCB has not a good ground plane.

i tested some type of antenna but i didn't get good results. in best situation, i got an RSSI about -3dBm on receiver side (my BEST result).
This result was obtained when the transmitter and receiver circuits were located at a distance less than 2 m and the transmitter output power was 100 mW.

i have about 7cm space for antenna and i don't know how to make a good antenna.

Any help is appreciated.

 

[FvM]

I don't understand your complains about RSSI, because -3dBm power at the receiver input is more than expectable according to Friis path loss formula with good antennas. But the RSSI measurement isn't that exact and near zone is also affecting the results.

A reasonable radiated power measurement would however use a calibrated receiver and antenna with known gain, e.g. a dipole or wide band measurement antenna. Related to λ/4 of 17 cm, 7 cm is clearly an electrical small antenna and you have to deal with the respective problems.

Texas Instruments has published various application notes related to 433 and 915 MHz small radio devices, including the antenna topic.

A document list can be found here Mesh and IP Networks - 6LoWPAN - CC1101 - TI.com

Particular interesting for the antenna problem are
**broken link removed**
https://www.ti.com/litv/pdf/swra161b

More specific about the TI antenna development kit, but instructive though
https://www.ti.com/litv/pdf/swra328

Finally, there's a large number of antenna text books and specific papers discussing small mobile antennas.

 

[hamed8419215]

First i should say that i am not an RF expert. I used RFM22B (si4432) RSSI register to measure RSSI. I read the register just after detecting preamble and i got a value about E0h that according to "RSSI vs Input Power" curve in SI4432 datasheet, the input power should be about -3dBm.

when i used a lambda/4 antenna, the RSSI register value increases up to EEh (that means about 4.5dBm input power). but i have a little space and i can not use lambda/4 antenna.

on the other side the RFM22B (with DIP package) has very little ground plane (also my PCB is single layer and has not any ground plane!). and all of the small size antenna that i saw need a ground plane.

with this situation (little ground plane and small size) which type of antenna is suitable for me?

 

[FvM]

The measurements are in the RSSI saturation range and no calibrated measurements anyway. But a difference between λ/4 and short antenna should be expected. A 7.5 dB delta (= about 4 dB gain loss for each antenna) won't be that bad.

It's right to think about ground plane. E.g. a λ/4 rod without a´groundplane is as good as no antenna. A monopole with a ground plane smaller than the antenna has to be analyzed as a kind of asymmetric dipole but hasn't much to do with the monpole idea that assumes an infinite ground plane.

In practice, the impedance of your actual antenna + ground plane combination has to measured and matched to the transceiver. Housing, battery, external cabling (if any) and "hand" effect of portable or mounting conditions of stationary devices have to considered. The trick is to utilize the existing ground for the antenna design.

An exact impedance measurement is a difficult task under these conditions, but possible. The achieved antenna matching should be validated by an ERP (effective radiated power) measurement of the final device, the measurement is required to check RF regulation compliance anyway.

In my opinion, the linked literature has useful hints for compact antenna design, many more can be found.

 

[hamed8419215]

can i reach to a lambda/4 antenna performance using a small size antenna?

 

[FvM]

Theoretically, an electrical small antenna can achieve the same performance, but with smaller bandwidth, which isn't an issue for 433 MHz devices.

In practice, the impedance matching network will add some losses (particularly by inductor series R) and the small Q and possibly low radiation resistance will make an exact matching difficult. So you should calculate a certain degree of mismatch. The high Q also involves a higher "hand sensitivity", de-tuning of the antenna due to external objects near to it. The undefined ground situation is the worst point in many cases, e.g. a change of the battery cable length or position considerably changes the antenna matching.

 

[hamed8419215]

but how can i calculate mismatch? i have not any professional measurement instruments. also i am not familiar in RF domain.
is there any way to design a practical small antenna without involving such measurement and heavy calculations?