Meandering Monopole PCB Antenna

[HackedFridgeMagnet]

Hi
From TI's Design Note   DN024  http://www.ti.com/lit/an/swra227e/swra227e.pdf

I attempting to use this concept for my own PCB antenna at 915 Mhz not 868 and definitely not the dual band version.

This is my first foray into a pcb antenna and so I was wondering if I had made any obvious mistakes. I have made a boat load of assumptions and no measurements and this I why I guess RF sometimes lies in the realm of black magic.

I basically cut down the length of L4 of the design not to 36mm from 38 and left the other measurements the same, my reasoning is I can cut the length a bit shorter if necessary in some sort of tuning process. I looked into the theory for the meandering antenna, it seems to be determined by total length but also heavily affected by the number of turns. The more turns the lower the resonant frequency.

Other queries  I have an 18mm nominally 50ohm trace to the impedance matching network. I chose the "Grounded Coplanar wave guide", instead of the "microstrip line" for no good reason. In the design note they have completely different dimensions. I don't know why mine would be any worse as it should be 50 ohms.

Then I have the impedance matching network I will probably just buy a range of caps and inductors to try.
I chose 0805 for the inductor and 0603 for the caps for no good reason. Are there any dielectrics or types that I should use or avoid.

My only piece of RF test gear is my $25 TI coupon bought spectrum analyser. My grasp of Smith Charts is somewhat diminished but I am keen to use them if I can somehow produce some measurements. I haven't got any SWR meter either. So it may be just RSSI strengths I can measure.

Even though I would like to get this right, I am obviously limited in test equipment and previously I had got some good range once with the antenna pulled off the coax. So I am ok with a partially compromised antenna system.


If anyone has a suggested plan of attack it would be appreciated.

ps. Green is bottom layer, Red-Brown is top layer- used as ground, Baby shit brown is both layers.

[HackedFridgeMagnet]

Bump, anyone done anything like this?

[Whuffo]

Bump, anyone done anything like this?

Nope - that's a tricky project and simulation is of very limited use. You're going to NEED to be able to measure reflected power; a SWR meter would work at a minimum. There's two impedance matches involved; one from the signal generator output to the antenna, and the one from the antenna to space. The meandering monopole is a compromise to reduce the amount of physical space the radiator uses, but it doesn't couple to space very well.

There's a bunch of tradeoffs in antenna design. Remember that a problem well defined is half solved and define it well before building. Do you need maximum range, minimum size, can you afford to uprate the output circuitry to handle a bad impedance match?

And ultimately, after you've done a careful design, you can build a prototype (or three) and then trim the length of the radiator for the lowest SWR. Then you can transfer that length to your design and go from there.

[HackedFridgeMagnet]

I take your point about a well defined problem.

Well my problem is sort of well defined in that, I have a box in which I can dedicate the top 40mm to being solely for the use of the antenna. And also about 100mm of width. Below this is are large flooded area at antenna ground potential.

Freq 915Mhz band, plenty of power for the required distance with a conventional antenna.

I will look into getting or making some sort of SWR meter.

[Whuffo]

I take your point about a well defined problem.

Well my problem is sort of well defined in that, I have a box in which I can dedicate the top 40mm to being solely for the use of the antenna. And also about 100mm of width. Below this is are large flooded area at antenna ground potential.

Freq 915Mhz band, plenty of power for the required distance with a conventional antenna.

I will look into getting or making some sort of SWR meter.

May as well get one; if you're going to be doing RF work it'll come in handy over and over again. Amateur radio suppliers usually have them available; they'll only be accurate up to their rated frequency, but you can go much higher if you're just tuning for minimum SWR (which is almost always the case).

And be careful about what kind of PCB material you use for this - it's not all the same and 915 Mhz is pushing it even for the good stuff.

[lgbeno]

Yes I think for a project like this, access to a SWR meter or better yet a VNA is essential.  I'm sure that your design will radiate but maybe not as optimally as one designed with proper tools.

Simulation is pretty important a too.  I've only used ADS but found this if your feeling adventurous http://www.qsl.net/4nec2/

[Whuffo]

Tuning antennas using field strength is an exercise in frustration & usually futility. Tuning for minimum SWR insures that the transmitter is properly coupled to the radiator; this will be a nice simple measurement that doesn't depend on distance, interference, etc. remaining stable over the tuning period.

Really - if you're building or tuning an antenna, you need a SWR meter.