2
\$\begingroup\$

I'm actually trying to focus an iBeacon that uses Bluetooth 4.0 LE. These transmit in the ISM band from 2.4 to 2.485 GHz. Normally these things radiate in a spherical, or half spherical pattern, but I'd like to restrict the pattern as much as possible to a beam - think a flashlight. If the device was on the ceiling, and 'beaming' straight down, then for the most part only receivers directly under it would detect the signals, and devices meters to the side would not see the signal directly. [I realize there might be scattering, nothing to be done about that.]

I googled around today, found nothing on this searching for 'Bluetooth signal shaping' (and focusing). Then I thought of that old trick to improve WIFI reception - put the antenna in a Pringles can.

If there is no obvious solution to this, what materials might be used as shielding to construct a signal barrier?

EDIT: After posting this, a list of related posts on the side showed up, and I saw the word 'cantenna'. Searching on Bluetooth cantenna' resulted in a few videos of people making such things to get directional signals. I also read that Bluetooth is using the same frequency (more or less) of WIFI so those cantennas should work too. Also, Anindo Ghosh posted an answer, he recommends a wire mesh to restrict the signal - so I've got a bunch of things to try now.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Use a dish antenna? Use a yagi? \$\endgroup\$
    – Andy aka
    Feb 5, 2015 at 19:13
  • \$\begingroup\$ @tomnexus - thanks - I'm EE but a few decades ago. Really, is there anything about this post that makes it appear less than clear and earnest? If a question garners two answers with 3 up votes each, how can the question possible be useless? [Unfortunately my StackOverflow score does me nothing on this forum.] \$\endgroup\$
    – David H
    Mar 24, 2015 at 19:06
  • \$\begingroup\$ Hi @DavidH. Did you solve this issue? I'm having basically the same issue using Estimote beacons and Raspberry Pis as receptors in the ceiling. I want to achieve the same thing, i.e. focusing the bluetooth signal so it's only detected by the RPi above. Did you try the cantennas? \$\endgroup\$
    – xtian777x
    Feb 25, 2020 at 4:53
  • \$\begingroup\$ @xtian777x Sorry, the project was cancelled and I never got to test these ideas out. \$\endgroup\$
    – David H
    Feb 28, 2020 at 20:52
  • \$\begingroup\$ @DavidH I tried the cantennas yesterday and they work amazingly good. That solved all my issues. Should I submit my experience as an answer here? \$\endgroup\$
    – xtian777x
    Feb 29, 2020 at 21:13

2 Answers 2

6
\$\begingroup\$

I assume you have a small beacon PCB that you haven't designed yourself, so you need to work with it as-is. if you're laying out the board, you have more options...

1. Restricting the area covered by a beacon
You're onto the right track here - mounting it above the supermarket aisle, or what have you, is probably best, then you can focus the signal into a small area. You might have to reduce the range or power of the signal somehow too. It's possible that only reducing the range would be sufficient.

2. Focussing the signal
Anido is right, you can focus the signal with some sort of antenna. As you probably don't have an antenna connector on the board, you will do best with a horn type antenna, like a pringles can, that is excited by the internal antenna on the beacon. A simple open-ended waveguide without a cone flare will give an 8-10 dBi antenna, which has about a 60-degree beam (at -3 dB), or maybe 90 degrees at -10 dB.

It must of course be metal, but the type of metal doesn't matter - aluminium foil stuck to cardboard, or a tin can. The metal must be continuous, without gaps, if you need a seam, have a good 1 inch / 2 cm overlap.

Look around the net for appropriate dimensions for the horn - the diameter matters more than the depth, so a soup-tin shape is OK. You could make a pyramidal horn out of cardboard too. No need to ground the transmitter to it - it's a reflector and concentrator. (You wouldn't ground a telescope mirror...?)

3. Measuring what you have done
It is fortunate that with BLE, you have an RSSI or signal strength indicator that is quite accurate. Invent a method of recording the signal strength, maybe using your phone or another device to record the result.
Mount the new transmitter-in-can on the ground, outdoors, facing up. You don't want any reflections from the ground to interfere with the signal. Move your phone or receiver over the front of it, perhaps 1 m away, and record the results. Keep your body far away from the test - use a stick to hold the phone/receiver, so you don't affect the results too much. The exact range isn't too important, it's the shape of the pattern you're measuring; it'll be almost the same 3m away as 1m away, just 10 dB weaker.
It's also important when measuring to worry about the polarisation of the signal, and the pattern of the receiving antenna. You can't change these, but you can hunt at each location for the phone orientation with the strongest signal.
From these results you will see if you're getting the desired focussing from the can-tenna that you've invented. If it's not working, try re-orienting the transmitter in the can, or moving it closer to or further from the back wall.

So in summary: Transmitter on the ground, outside. Move phone around slowly above it, to get an idea of the shape of the radiation pattern. At each position, twist the phone around to get maximum signal. Adjust beacon position in tin can, until you like the pattern.

4. Reducing the range
You'll probably find that the signal spills over where you want it to go, even (especially) with the antenna focussing it. Time to reduce the signal power. You might be able to do this in software, on the BLE beacon, you might have to resort to some hardware modifications. Soldering a resistor in parallel with the antenna on the beacon will reduce its range, 10 ohms might be a good starting point. If you're not that brave, you could try wrapping it slowly in aluminium foil (after insulating the board, of course) to reduce the amount of radiation by creating a mismatch in the antenna. Even just placing it flat on the base of the tin, or right in one corner, might do the trick.

5. Experiment! Apart from the can dimensions, everything else is negotiable. Experiment! Remember that you can always measure the signal strength with your receiving device, to compare before/after, and see what you've done.

\$\endgroup\$
3
\$\begingroup\$

Assuming the OP has access to an external antenna connector on the device, one option is to use a Yagi-Uda directional antenna tuned for the 2.4 GHz band. This question touches upon a similar Bluetooth problem statement: adding a yagi antenna to a bluetooth dongle.

If an integrated device needs to be designed with directional beamforming, this can also be achieved using microstrip antennas in an array formation. This is perhaps not for the faint of heart, but this paper and others like it will provide some insight.

Finally, if signal loss is not a concern, a lossy attenuation of the portion of the signal that diverges far from the designed axis can be achieved: A hollow tube of fine metal mesh (fine chicken mesh might serve), grounded, projecting from the actual bluetooth device outward along the desired axis for a bit more than 12.5 cm (the approximate wavelength of the 2.4 GHz signal in air).

\$\endgroup\$
2
  • \$\begingroup\$ The off-the-shelf units have an on-circuit-board antenna, buried in an interior layer of a multi-layer board. That said other iBeacons might have an exterior antenna. The 5 inches or so of fine mesh sounds like a perfect idea, at least for now. The units are battery powered - should I ground to the battery '-' terminal, or to an electrical ground? I'll add an additional comment when I get the wire mesh. \$\endgroup\$
    – David H
    Feb 5, 2015 at 19:48
  • 2
    \$\begingroup\$ Ground to the electrical ground that the antenna references. I have no way of guessing whether the battery negative corresponds to that, it need not. \$\endgroup\$ Feb 5, 2015 at 19:50

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.