I used to install a lot of in-car UHF radios & hands-free cell phone kits. I mostly worked on trucks & commercial utility vehicles fitted with bull bars, upon which I would mount a big, bulky, high-gain antenna. Alternatively, a bracket could be mounted directly to the chassis under the bonnet (which often meant cutting/drilling). The coaxial cable would run from the new device inside the cab, through the engine-bay, and out to the base of the antenna.

enter image description here enter image description here enter image description here

But every now and then I'd get some fancy executive with a shiny new luxury sedan, and I'd use a relatively discrete glass-mount antenna. The cable runs under the dash and up the pillar (under the upholstery or plastic trim) on the passenger side, and pops out near the top of the windscreen. The cable screws into a little black box/panel, one side of which adheres directly to the inside surface of the glass. At the base of the actual antenna mast is a similar adhesive panel that mounts on the exterior surface of the glass, directly on top of the first.

enter image description here enter image description here enter image description here

I never fully understood how or why it worked, but essentially the signal was able to flow right through the glass. My question is: Can this same technique be adopted for 2.4GHz and/or 5GHz WiFi antennas?

  • \$\begingroup\$ In theory UHF works from 300Mhz to 3Ghz so it should work for lower F antennas. However you should check the bandwith. But if you can test it I would like to see the results. I find this post very interesting. As for the way it works is throug inductance. \$\endgroup\$ – Dimitri Jan 26 '18 at 12:21
  • \$\begingroup\$ Probably transfers energy to the outside part using inductance (see contact-less/inductive charging). Then the RF is amplified outside and send back through the glass again. They might even use the same coils but at different frequencies. \$\endgroup\$ – Oldfart Jan 26 '18 at 12:35
  • 4
    \$\begingroup\$ "new luxury sedan" and "no built-in hands-free capability" is a set of mutually exclusive conditions nowadays... \$\endgroup\$ – ratchet freak Jan 26 '18 at 16:46
  • 1
    \$\begingroup\$ @Dimitri 300MHz - 1GHz, according to the IEEE; or 300MHz - 3GHz, according to the ITU. But I was actually referring to the 477MHz UHF Citizens Band used in Australia & New Zealand (476.4250 MHz - 477.4125 MHz; colloquially known simply as UHF). I thought this was a common standard throughout Europe & North America, but apparently not. \$\endgroup\$ – voices Jan 28 '18 at 1:46
  • \$\begingroup\$ @tjt263 Ok. Now you have a problem then. You can't simply use the antena used for ITU-R in case of WIFI Frequencies. You should make check this foto for example: robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/…. As you can see different length for different Frequencies. I won't lie to you I have no idea if it would work if you modify the antena you have, but that could be a good experiment. Check in google "antena length calculate". You will see a few useful formulas. Hope this helps friend. \$\endgroup\$ – Dimitri Jan 29 '18 at 7:53

Parallel capacitor plates of 25 mm by 25 mm seperated by 4mm of glass with a relative permittivity of 4 would give a coupling capacitance of about 5 pF. That capacitance is in series with an antenna signal and at 2.5 GHz, would act as a blocking impedance of about 13 ohms so, it's feasible it could be used without disrupting the VSWR too much.

enter image description here

  • k is the relative permittivity of the glass.
  • \$\epsilon_0\$ is the absolute permittivity of free space (8.854 pF/m)

The 13 ohms series blocking impedance could be tuned out by a small value of series inductance. I would expect it to work best when located at the edge of the window (and close to the car body) because the antenna type is a monopole and it needs some form of local ground plane to be most effective. In other words, using a capacitive connection requires that the base of the antenna is one plate of the capacitor.

I'm not going to rule out coils that can couple via the glass window but, at 2.5 GHz, these might be starting to become lossier than a capacitive coupler.

  • 3
    \$\begingroup\$ There are also other advantages to having the antenna at the edge of the windscreen rather than right in the middle ;-) \$\endgroup\$ – psmears Jan 26 '18 at 16:48
  • \$\begingroup\$ @psmears what are they (other than the one I listed)? Don't be shy on this! We're all here to learn something. \$\endgroup\$ – Andy aka Jan 26 '18 at 16:50
  • 7
    \$\begingroup\$ Sorry, that was intended as a joke - I just meant that having the antenna at the edge is also better for being able to see where you're driving :-) \$\endgroup\$ – psmears Jan 26 '18 at 16:56
  • \$\begingroup\$ Thanks for the information. It's a bit out of my league at this stage, but I'm wrapping my head around it.. So anyway, is it normal to have a capacitor, (or pair of inductors) immediately before (and in series with) an antenna? Or is it simply a means of traversing the glass? It seems like the kind of thing that would have a significant affect on the behaviour of the main circuit. \$\endgroup\$ – voices Jan 29 '18 at 8:54
  • \$\begingroup\$ If you look into matching networks between chips and wifi antennas it's not uncommon to see small pF value capacitors in series. For wifi frequencies I've never seen a pair of coupled inductors and I suspect that is because the frequency is too high to get effective magnetic coupling like this. \$\endgroup\$ – Andy aka Jan 29 '18 at 10:03

I never fully understood how or why it worked

Well, it's not magic ;-)

Actually it can either be done

magnetically using coupled inductors. This is like a transformer without a magnetic core. Wireless charging as used in some mobile phones uses the same principle. Basically, one coil creates a magnetic field from an electrical signal which is then picked up by a second coil (at the other side of the insulator, that can be any insulator including air or glass). The second coil turns the magnetic field back into an electrical signal.


electrically using a capacitor like structure. A capacitor consists of two electrically conductive plates with an insulator (that can be any insulator including air or glass) in between. A capacitor is a low impedance (does not form an obstacle) for high frequency signals.

For lower frequencies up to 200 MHz or so, I expect the magnetic coupling method to be used. For low frequencies a very large capacitor (for the electric coupling) would be needed to be efficient.

For high frequencies above 200 MHz, so that includes WiFi signals, I expect the electrical coupling method to be used. High frequencies cannot travel through large coils making the magnetic coupling method difficult.

  • \$\begingroup\$ So the glass is basically a dielectric plate? What about the ground wire/braid/conductor; how do both sides of the circuit traverse the glass through a single point/component/capacitor? \$\endgroup\$ – voices Jan 26 '18 at 15:41
  • \$\begingroup\$ Correct, the glass would be the capacitor's dielectric material. There will be no ground at the antenna, the ground will be the car's chassis I guess. Also you don't always need a ground, an antenna can work with only one connection. \$\endgroup\$ – Bimpelrekkie Jan 26 '18 at 16:31
  • \$\begingroup\$ Is that right? That's weird then that we continue to run relatively expensive coax with the negative braided sheath to the outer conductor of all our antenna connectors. Most of them even come with special washers that bite into their mounting surface in order to ensure a strong path to ground. \$\endgroup\$ – voices Jan 26 '18 at 16:53
  • 1
    \$\begingroup\$ @tjt263 it is right. If you would run single wire, it would act as a longer antenna, not as a way to transmit signal from the antenna. So if it's really, really short, ie connected directly to the PCB, it works. But that's a matter for another question. \$\endgroup\$ – Mołot Jan 26 '18 at 20:22
  • \$\begingroup\$ So basically, it's a 6 foot antenna; spanning from the centre console to the top of the windscreen? Something about that doesn't seem right. \$\endgroup\$ – voices Jan 27 '18 at 8:17

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.