Small PCB antenna 6

[FriendlyBen]

Hi everybody,
I want to radiate out of a 1" X 1½" board at 434 and 868 MHz ISM band. I am now 27 dB lower with that small loop around the board pcb antenna in both frequencies than a bigger board which have an helical wire antenna. Both project have the same short range transmitter circuit (arround 5 dBm output). Both circuit was best tuned. Yes, I know that I will never reach this 27 dB gap, I am not a dreamer. But at least, is anybody know how can I gain some dB ? Any experiences on this issue is welcome !

Thanks
FriendlyBen

 

[Comcokid]

Are you trying to do both bands with the same antenna?

Getting good radiation out of such a small antenna will be difficult. For ideas, I suggest you try searching for "electrically small antennas" which has given me pointers for the 915 MHz US ISM band.

You can shrink an antenna size by dielectrically loading it. That is the method used for some PCB antennas sold by several companies such as the following one.

http://www.linxtechnologies.com

 

[FriendlyBen]

Hi Comcokid,

I am doing both frequencies on the same antenna (I must use the same pcb layout), both have a matching circuit and both shown about the same attenuation. Thank you to remember me Linx tech. I already try somes samples, it's very good for the size. They play with the feed point location to get the best match. They use FR-4, I don't think they dope the dielectric and I must stay with the FR-4 too. Overall performances and radiation pattern are quite good also. Maybe it's a good path to follow...

Any other good idea ?

Thanks !
FriendlyBen

 

[Comcokid]

FrendlyBen,
Remember, just because you get a good match into your tiny antenna doesn't mean it will radiate efficiently. Experiment with different antenna layouts.

Since you're doing both bands with one layout, you could tap at a different point on the antenna for each assembly with the presence/absence of a SMT resistor or capacitor.

Also, don't overlook your enclosure. You can experiment with a strip of metallized tape or patterned conductor or wire inside whatever enclosure/case you're using. Your PCB antenna will capacitively couple to this conductor, and re-radiate, hopefully more effectively.

I opened up my 'antenna' file, and found the following articles that may be of some help.
1) Techniques for Determining Small Antenna Gain, RF Design Magazine, May 2003, pp62-68
2) Small, 915-MHz antenna beats monopole, EDN Magazine , "Design Ideas, May 15, 2003, pp73-74 (uses an antenna about your size, maybe you could rescale for your frequency)
3) Designing dual-band internal antennas, EDN Magazine, Nov 8, 2001, pp99-103
4) Antennas for Low Power Applications, Application note downloaded from the RFM website.
5) Small Loop antennas, Application note downloaded from Nordic VLSI website. (they have other good app notes as well.)
6) MICRF001 Antenna Design Tutorial, Application Note 23, downloaded from the Micrel web site.
7) Matching Small Loop Antennas to rfPIC Devices, Application Note AN831 , downloaded from the Microchip web site.

 

[logbook]

I am not sure if you are driving this antenna as a loop or as a short monopole. If it is a short monopole then the ground plane is the key to getting more emission. The ground plane is obviously not under the signal wire, but is the other half of the antenna. When we were working with a low power radio module at 418MHz (?) the ground plane increased the effective range considerably. (This was with a short helical antenna.)

 

[FriendlyBen]

Lot of material to chew on, great !

Thanks !
FriendlyBen

 

[kromvas]

You should try F-antenna.

 

[FriendlyBen]

Hi everyone

Right now I got 7 dB improve with a PIFA/meanderline like antenna at 433 MHz. DOUBLE THE RANGE ! great ! let's try 868 MHz for the next weeks...to be continue...

Thank's to everybody ideas!

Thanks !
FriendlyBen

 

[FriendlyBen]

Just try the 868 MHz, already 17 dB higher just by tuning and with the same PCB layout ! More than expected ! So great !

Thanks !
FriendlyBen

 

[RFtweaker]

Hi everyone,

I'm also starting to work on PIFA and/or F-antenna but I do not know how to compute the dimensions of these antennas. Do u have some papers about it? I can see you have some interesting results...

Thanks in advance

 

[FriendlyBen]

HI RFtweaker,
I am an RF technician, the way I use to do it is trial and error. I moved my ground and feed point until i get 50 ohms on the network analyser. Also I cut the length of the antenna until I get the max power out. It's an almost blind process and time consumming also. Right now I try to increase the power by changing the layout to gain some more dB's. I have many days of fun in front of me...
Sure some people can provide you some computation formula, I suggest you to start a new tread, you may better catch the attention of some engineer that is not involved in this tread.

Good luck !

Thanks !
FriendlyBen

 

[FrankTCS]

FriendlyBen,

Can you tell me about your progress on the 433 MHz antenna. Did you get a working design with good gain?

Thx..

f

 

[FriendlyBen]

Hi FrankTCS,
At 433MHz I got a working design, meaning in production.
At 868MHz the solder mask make me face a big self oscillation problem.
The answer to your question is subject to interpretation: "Did I get a good gain" for a personnal point of view I would say, not enough now ! But for the corporate they were satisfied with the range. All depend on the expected range. For sure a so small pcb antenna will never radiate out as a bigger external one ! I think that there is still some place for improvement. I am working now on a different layout to get better performance. It is a long process and stability must be the first concern. It is not only a question of getting more dB, but more stable and easy to produce dB's.
Hope this help !

Thanks !
FriendlyBen

 

[VEBill]

Make sure that you keep the two aspects of gain, loss and directivity, separate.

Loss, in general, is just plain bad. Antennas that are small in proportion to the wavelength tend to be lossy, often due to Lo-Z, resultant high currents and voltage drops. If you can improve conductivity, then you can reduce the loss.

Directivity can provide gain, but certain directions benefit to the exclusion of other directions. For many applications, directivity is also bad (if you can't control the aiming of the device).

There are many applications where the goal is to minimize loss (which improves gain) and also minimize directivity (which reduces peak gain).

In other words, be careful about seeking high gain if your application requires a wide coverage pattern and you can't control the aiming of the device.

 

[FriendlyBen]

VE1BLL,
I agree with you !
Actually for me I have a fair good radiation pattern and I already reach my electric 1/4 Lambda lenght by layout strategy (Printed inverted F folded monopole). I think that the corporate already found that using the same pcb for both 433MHz and 868MHz for printed antenna is more part of the utopy than reality.
I am facing two chalenge now, loss in heat rather than radiation and keeping the plane big enough. Since I added solder mask make loosing gain, my next steps will be to avoid solder mask over the antenna to avoid the dielectric effect on it and also take care of the size of the ground plane. That's a very big and fun chalenge in miniaturisation ;-)

Thanks !
FriendlyBen

 

[rfsd77]

Does anyone have material for desiging an printed circuit board inverted F antenna, I know someone else had asked but saw no update.

 

[FriendlyBen]

This is a paper that help me understand F antenna

http://www.loe.ee.upatras.gr/EM/Jou...on a PCMCIA Card for the 2.4 GHz ISM band.pdf

Hope this may help !

Thanks !
FriendlyBen

 

[cgorney]

How exactly do you trace a PCB antenna? Is it simply a trace that deadends or do you create a loop on the board? I need to receive a 418MHz signal (using a Linx receiver) and I need to incorporate the antenna onto the board. Thanks.

 

[FriendlyBen]

Cgorney,

I don't know about your experience in RF but there is many ways to trace an antenna on a PCB, all depend on you space constraint, desired gain, range, etc. If you don't have RF equipment like a network analyzer you should consider buying a SMD antenna like

http://www.linxtechnologies.com

offer. There are many parameters to consider and it is not straightforward. Here we have both loop and folded monopole (deadened) on our products and both behave differently and are being calculated differently. You can go on

http://www.rfcafe.com

or

http://www.rfglobalnet.com

to get some information and white paper that may help you in your design.

Good luck


Thanks !
FriendlyBen

 

[Comcokid]

For an antenna, you could check the types and sources I posted earlier. Some of these can be realized as a PCB trace.

In addition, an inverted F antenna could be used, although it is getting rather large at 418MHz. Check:

http://www.poynting.co.za/tech_training/ifa.shtml

What you do is reinterpret what is shown at this web site to be retangular traces that are "beyond the ground plane edge" rather than "above the ground plane"