I just built a Bluetooth speaker using a Sure Electronics AA-AB41136 Bluetooth 4.0 Audio Receiver Board. The reception was not great since it is a PCB antenna contained in my speaker box which is made out of 1/2" thick plywood. It seems a lot of people are extending these antennas using RG-58 wire and I just happened to have some laying around. From what I read the length of the antenna can be from 28.8mm to 32mm for a 1/4 wavelength on the 2.4 GHz spectrum. The length of the antenna would start after the shielded portion of the RG-58 ends.

I cut the trace on the PCB antenna leaving about 5mm and attaching about a 4mm length of the bare wire to it. I then fed the shielded portion up to the inside top of the speaker and cut back the shielded portion and fed 43.7mm (31mm plus the 1/2" going through the wood) out the top.

After that long winded explanation my question is actually quite simple. Using 31mm as the antenna length does the 1/2" and/or 5mm on the trace end need to be factored in to the equation? Should I cut my existing 43.7mm down or not so the total unshielded portion going through the wood and out the top would be 31mm? Or if I must included the 5mm on the PCB end would it be 26mm through the wood and out the top? Or some other variation of something I am missing?

enter image description here enter image description here

  1. Those measurements are for wire elements in free air; not with the insulation still covering the wire. The value of the insulation will affect the surface electrical speed of the wire, and thus change the length you'd need (easiest fix is to cut the insulation back).

  2. The length of wire that is passing through the box, with the shield cut off must be counted towards your antenna's effective length.

  3. The exposed few mm of trace & the exposed length of coax center conductor you soldered to it should be electrically shielded to minimize interference (both caused by the antenna & interfering with it).

  4. For a quite effective, cut-coax antenna with better reception (2-3 dBi gain), look up a "Bazooka" Dipole Antenna. It's quite easy to make (I've personally made more than a dozen of them, for various bands) & has decent performance without being too "fiddly" when it comes to getting the measurements exactly perfect.
  • \$\begingroup\$ Thanks for such a thorough explanation. I was basing my design off of this electroschematics.com/712/24ghz-coaxial-antenna which I believe is a dipole but I had no idea where to solder the shielding to on the PCB side. The reception I am getting now is quite improved at least 100 feet line of site. I will try cutting it though and see what that does. \$\endgroup\$
    – Hartman9
    Jun 13 '16 at 16:19
  • \$\begingroup\$ Instead of shielding the PCB connection, you could try and match the PCB to the cable to minimize loss. It sounds like the PCB design isn't yours, but this is pretty common. digikey.com/product-search/en/connectors-interconnects/… \$\endgroup\$
    – Andrew W.
    Jun 13 '16 at 17:22
  • \$\begingroup\$ @Hartman9 I just skimmed through the instructable you linked. That design is for a 1/4-wave "whip" style antenna with ground "plane" radials. Notice that he paid special attention (mention) to soldering the 4 bundles of shielding braid (to stiffen them into 'solid' radials), then used gobs of hot glue to hold them in position. That antenna design is very similar to the 1/4-wave whip antennae used on military & police vehicles (and some HAM radio operators' rides too), in which the (usually) metal body of the vehicle forms the ground 'plane,' the same as the 'radials' in that design. \$\endgroup\$ Jun 14 '16 at 8:45
  • \$\begingroup\$ While that antenna design is quite valid (although his measurements could use a bit of tweaking due to the reduced electrical velocity inside that insulation -- I think for teflon insulation it's 75/100 speed/length iirc), I prefer the bazooka-style dipole like I linked in the answer due to its being a bit more space-conscious (no protruding radials), potentially has slightly higher gain, and is (in my opinion) a bit more rugged, since the inverted braid sleeve is harder to deform that protruding radials. \$\endgroup\$ Jun 14 '16 at 8:52
  • \$\begingroup\$ As for grounding the shielding; you can generally "get away with" soldering it directly to any power (or 'main') ground pour on the pcb, though if you have a schematic, 'signal ground' would be best. - - IMPORTANT Failing to solder the shield braid to ground on your pcb will result in the braid becoming an 'active' part of the antenna. This will cause the braid to receive/send signals in unpredictable ways, and can cause some very 'interesting' effects (which are usually best avoided). \$\endgroup\$ Jun 14 '16 at 8:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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