7

Coaxial cables with RP-SMA connectors are often used as extensions for external antennas / with software defined radios.

What would happen if I were to connect two access points (wifi chips, via their antenna ports) using a coaxial cable? Would they be able to communicate? What should I take into account to not damage the devices?

I am trying to extract the angle of arrival from the channel state information reported by the wifi chip. Therefore I need to remove the phase shift introduced by the chips design. I can do this by measuring it, for that I need to receive a signal with the same phase at both antennas though. This is pretty hard in a multipath environment therefore the only solution I can think of at the moment is connecting both 802.11 chips via cable.

Geosearchef
  • 173
  • 4

4 Answers4

7

Sure, just attenuate it (50-60dB). You don't want to saturate the receiver. As for collisions, they happen on air too (at different power levels).

If you have a MIMO NIC with multiple SMAs, you can even play with the phase a bit.

Lior Bilia
  • 7,282
  • 1
  • 20
  • 30
0

Learn what your hardware can tolerate and it's specifications like is it self-terminating? how large a signal input can it take without damage?

As to your experimentation objective, you could use different coax lengths to control received signal phase.

FYI, Legacy 802 series ethernet bus using coax (thinnet, thicknet) did require 50 ohm termination at each end. If you are experimenting with components intended for connection to antennas (that's my impression), that's a totally different animal and I agree that attenuation may be required to protect the receivers.

r.f.geek
  • 1
-2

Probably not a good idea. Not only would matching be a nightmare, but what happens when both radios decide to transmit at the same time?

See my answer here for a bit more info.

Hitek
  • 852
  • 4
  • 10
  • Thanks for the love! OP asked about connecting the SMA connectors directly to each other, which is not a good idea. I agree that I should have added the use of an attenuator, or rather, not expanded the answer in that direction at all since it doesn't answer the original question. I will remove the part of the answer that is out of scope. – Hitek Aug 29 '19 at 03:42
-3

This should work fine, you just need to terminate the SMA cable at both ends with 50 Ohm termination resistors on SMA T adapters.
You don't need to introduce attenuators at all.

Jack Creasey
  • 21,428
  • 2
  • 15
  • 29
  • 2
    All that kind of "termination" will do is create an impedance *mismatch*. But probably not enough of one to save the frontend from blowing up when it gets 60-70dB more power than it's expecting. So why not use an attenuator, again? – hobbs Aug 29 '19 at 05:16
  • @hobbs, then you don't understand termination..... The output of the RF amplifier is conventionally 50 Ohm, and the cable is 50 Ohm, and you need to terminate BOTH ends of the cable. – Jack Creasey Aug 29 '19 at 15:01
  • The receiver is *also* 50 ohm, because it's designed to work with the same antenna/cable system as the amplifier, so the business with the tee and terminator will show the cable a 25-ohm load. – hobbs Aug 29 '19 at 15:05
  • @hobbs ….so clearly you do not understand RF cable termination at all!!!!! https://en.wikipedia.org/wiki/Electrical_termination ….we are NOT dealing with Ohms law here. – Jack Creasey Aug 29 '19 at 15:24
  • https://en.wikipedia.org/wiki/Electrical_impedance#Parallel_combination try learning the basics of the field you're trying to school people in. Impedances in parallel transform just like resistances in parallel. – hobbs Aug 29 '19 at 15:46
  • @hobbs nothing wrong with the link you posted ...BUT, it deals with a complex (impedance) and not just a resistance. For RF cable you MUST terminate at BOTH ends. If you do not then one end of the cable will be open circuit and you will get 100% reflection from that end for the signal (which takes time to pass along the cable). Time for you to do some reading I'm afraid. My apologies if you feel insulted ….but don't make comments such as "All that kind of "termination" will do is create an impedance mismatch." when you don't understand the technicalities. – Jack Creasey Aug 29 '19 at 15:53