0
\$\begingroup\$

I am building a weather balloon sensor baord with a LORA 868 MHz radio module (RFM95x). The sensor will rise up to 30km and should transmit the full time.

I am using a PCB trace antenna with ground plane. For ground applications it is fine to have the antenna on the upper side of the board.

My question A is if I should flip the board upside down to get the ground plane out of the way and have a clear view to the basestation?

Question B is if I flip the antenna (rotate the board), what happens to the polarizastion? Do I have to consider this in the trace (or basestation antenna)? Is this type of PCB antenna the way to go, or should I use an external (dipole) antenna?

Thank you for the help!

868 MHz PCB antenna like this:

enter image description here

\$\endgroup\$
5
  • \$\begingroup\$ Use the very best antenna you can afford - have you analysed the path loss at 30 km? \$\endgroup\$
    – Andy aka
    Commented Dec 10, 2019 at 11:55
  • \$\begingroup\$ Yeah ... unfortunately it has to be cheap, so I'd really like to stick to something simple and easy to manufacture. The current record of 868 LoRaWAN is at around 700km in europe. thethingsnetwork.org/article/lorawan-distance-world-record So I hope, considering the good coverage of ground stations to have a connection all the time the balloon rises. \$\endgroup\$
    – Jan
    Commented Dec 10, 2019 at 12:37
  • \$\begingroup\$ "should transmit the full time" it had better not! Review the regulations in your location, 868 band has duty cycle limits, additionally an airborne transmitter sending more often than it needs to is quite annoying over a large geographic area to other users of a shared band. \$\endgroup\$
    – Chris Stratton
    Commented Dec 10, 2019 at 15:02
  • \$\begingroup\$ This is not however a particularly suited antenna - it is a compromise for small gadgets. On a balloon you have plenty of space and could consider a full sized dipole or whip as might be used on a gateway - your enemy is weight not size. And your receiver in the critical case is more likely on the horizon than "down" \$\endgroup\$
    – Chris Stratton
    Commented Dec 10, 2019 at 15:12
  • \$\begingroup\$ Thanks Chris, this helps a lot. With "should transmit the full time" I meant "should be able transmit the full time". Of course I will respect the 1% rule. A wire antenna might actually be the better solution. \$\endgroup\$
    – Jan
    Commented Dec 10, 2019 at 17:40

1 Answer 1

Reset to default
1
\$\begingroup\$

Without knowing what other analysis you have done, I would recommend doing what you suggested and flip the PCB so that the ground/reference plane is "up" and the radiator is "down".

While you will probably get some performance with the ground plane down, that will be because of RF leakage around the edge of your finite ground plane, and maybe from reflections from objects above the antenna. If the ground plane were infinite, you would get absolutely nothing (or close to it) on the ground plane side of the antenna.

Did you or anyone else simulate the antenna performance using an E&M tool such as CST Microwave or HFSS? There are other tools out there, but those are the two that we use for antenna element pattern analysis and design.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for the answer! Currently we have no simulations for that type but it is proven to work on ground applications. Maybe I'll just give it a try and see how far we get. I'm not really in depth with rf circits, but polraization should not be affected, right? \$\endgroup\$
    – Jan
    Commented Dec 10, 2019 at 11:51
  • \$\begingroup\$ Jan, if the ground plane is down, then the antenna can only emit up. Polarization at that point doesn't matter. \$\endgroup\$
    – Marcus Müller
    Commented Dec 10, 2019 at 13:38
  • \$\begingroup\$ This seems based on a misunderstanding of the board, there is no ground on the PCB in the area of such an antenna, rather the ground plane is in the area of the PCB containing the components. So it while it may affect the radiation pattern somewhat it will not function as a shield. Nor is the receiver likely to be directly below the transmitter for long. \$\endgroup\$
    – Chris Stratton
    Commented Dec 10, 2019 at 15:06
  • \$\begingroup\$ I was going from this from the OP: "I am using a PCB trace antenna with ground plane. For ground applications it is fine to have the antenna on the upper side of the board" \$\endgroup\$
    – SteveSh
    Commented Dec 10, 2019 at 22:31
  • \$\begingroup\$ Yes. Sorry. Board view is not clear. Actually both sides of the board are on GND. The line below the antenna is the border of that. It is just disabled in this view. \$\endgroup\$
    – Jan
    Commented Dec 11, 2019 at 23:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.