0
\$\begingroup\$

Is there any efficient way to make an antenna radiate to the outside from inside a close metalic enclosure? I cannot drill any holes in it, nor place the antenna on the outside, nor change the material type.

I know that the metal acts as a shield and reflector for the radio waves. I was thinking on using a directive antenna and hope that the radio waves find their way out/in somehow, maybe through some open edges. This seems like some sort of impossible task, you have any alternatives to it?

I'm working over the GSM band.

Regards

\$\endgroup\$
9
  • 2
    \$\begingroup\$ You seem to have eliminated all the possible options that would allow it to work. \$\endgroup\$
    – user16324
    Jul 27, 2017 at 13:40
  • \$\begingroup\$ It's a modem with the max power of 1W into the antenna. The device is powered to the domestic grid, so power may not be a problem it is a solution The used metal is steel \$\endgroup\$ Jul 27, 2017 at 13:42
  • \$\begingroup\$ Should be deployed on already deployed matalic enclosures. Wouldn't be viable. And the antenna shouldn't be seen from the outside to prevent it from being stolen \$\endgroup\$ Jul 27, 2017 at 13:43
  • \$\begingroup\$ You can't drink from the bottle until the stopper is removed. \$\endgroup\$
    – Andy aka
    Jul 27, 2017 at 13:45
  • 1
    \$\begingroup\$ Don't you need two-way operation for GSM? Even if you can somehow get a signal out of your enclosure you still need to receive from the other end, and that's even less likely. \$\endgroup\$
    – Finbarr
    Jul 27, 2017 at 14:29

1 Answer 1

4
\$\begingroup\$

As you say, the enclosure will block radio waves emitted inside of it.

People often think that a shield should block all radio waves. Technically it does, but in practice it doesn't because the shield itself re-radiates some of the power. You may be able to use this to your advantage.

If one of the dimensions of the box is a multiple of ½ wavelengths, then use the box as your antenna. Connect the two antenna feed lines to opposite points of the box on the inside. Unless you have sophisticated antenna modeling software, you'll need to experiment with a field strength analyzer to see what connection points work best.

\$\endgroup\$
4
  • \$\begingroup\$ Wonder how efficient that would be. But may be something to consider and test, thank you for the idea. \$\endgroup\$ Jul 27, 2017 at 14:05
  • 1
    \$\begingroup\$ Lets just hope the other systems present in the box are not susceptible to bursts of RF. Anyway, GSM modems (at least the several I've used) are generally able to test their antennas' efficiency either with built-in test modes or by manually polling signal levels with the AT command-set (or other library). In the absence of exotic test gear or software this might be enough to fine tune the placement of the internals. \$\endgroup\$
    – user98663
    Jul 27, 2017 at 14:09
  • \$\begingroup\$ @Woss: Susceptibility of other parts of the circuit to RF is a valid concern. It might be necessary to have a secondary sheild inside the box to protect other sensitive circuitry. Of course being extra careful with good grounding is the first thing to do. \$\endgroup\$ Jul 27, 2017 at 15:30
  • \$\begingroup\$ @OlinLathrop, I'm no expert in Electronics or RF but I have had some highly interesting involvement in EMC testing of our products. Some things that sprang to mind with the OP's question were: "accidental slots in the enclosure that might be exploited for fun and profit" and "there's a danger of adding either harmful interference, or harmful Conducted Emissions". Two sides of the same coin perhaps. I think every EE (qualified or not) should spend some time in an anechoic RF testing facility just to get some grounding (hah!) in RF science. \$\endgroup\$
    – user98663
    Jul 27, 2017 at 19:36

Your Answer

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

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