Author Topic: Connecting RF modules 'directly' through cable  (Read 2937 times)

0 Members and 1 Guest are viewing this topic.

Offline martinmunkTopic starter

  • Newbie
  • Posts: 2
  • Country: dk
Connecting RF modules 'directly' through cable
« on: January 25, 2017, 02:35:14 pm »
Hey fellow geeks!

I finally registered after being passive for quite a few years in here!

Anyway, i was wondering if you could help me out with my limited knowledge  ???

My client do software for media streaming over Wifi, and I have created an elaborate test/QA setup for all of their supported HW.

So-far, however, I have been doing testing over ethernet with DummyNet (network simulation) profiles for having a repeatable test situation.

I would like to extend this to testing over wifi as well and was wondering if some kind of "box" or "switch" exists that i could connect 10 or so wifi modules to via coaxial cable and they would interact as if it was in a perfect faraday cage?

In fact, I think i have seen something like this in a video about the production of Digi's XBee modules, but really cant find it now.

I don't know if this kind of gear exists off-the-shelf and i really don't know what to search for  :-//

Please post your thoughts!

- Martin Munk
 

Offline ConKbot

  • Super Contributor
  • ***
  • Posts: 1401
Re: Connecting RF modules 'directly' through cable
« Reply #1 on: January 25, 2017, 05:28:58 pm »
You could use a 10 way splitter/combiner with the sum port left either open or shorted. Power that comes in will bounce off of the open/short and get split back to all the ports. You can even put an attenuator on the sum port (again, open or short the output of it) but you will be limited in how much you can attenuate the signal down by the isolation on the splitter/combiner, so you might need individual attenuator for each module also.
 

Offline NavyBOFH

  • Contributor
  • Posts: 32
  • Country: us
Re: Connecting RF modules 'directly' through cable
« Reply #2 on: January 25, 2017, 05:31:24 pm »
Someone else might chime in here - but generally RF out and RF in are two wildly different measurements. For example, a Wifi adapter might put our 100mw but the receiver will be able to pick up (let's just say...) -60dBm - which will be microwatts of power relatively. If you took a Wifi adapter and wired it directly to another one, you'll more than likely blow out the other device, if not both.

As for a "perfect world" type testing for Wifi, there isn't much of one unless you are in a Faraday cage or other RF-insulated enclosure in which you can measure TX, RX, and throughput in a "perfect" setting.

Maybe something can be wired up using the access points, a directional coupler or attenuator, and to another unit - but not likely since most Wifi standards are simultaneous transmit and receive over the same antenna - so attenuating will be difficult.

Best bet (in my mind) - just follow manufacturers recommendations for their products and assume for worst-case scenario. In the broadcasting field - we are heavily anti-wireless and everything on our networks (at transmitter sites) has a scheme in which all streaming of AES (FM) and SMTP (DTV) has priority over all other IP traffic all the way down to IP doesn't get multiplexed off the DS3 circuit until AFTER all video/audio multiplexing.
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5057
  • Country: si
Re: Connecting RF modules 'directly' through cable
« Reply #3 on: January 25, 2017, 06:58:07 pm »
You can easily reduce the power coming back out of that splitter/combiner by adding a attenuator on it. In almost all cases they are not directional so it doesn't matter if its transmitting out or its getting power back trough it.

So put for example 30dB attenuators on each port of your splitter and the transmitting power gets cut down by 30dB as it goes in, then maybe the splitter looses 15dB and gets attenuated by 30dB again as it comes back out to the receiving device in total 75dB of loss. And that's something that is probably a reasonable signal strength seen in every day use.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17498
  • Country: us
  • DavidH
Re: Connecting RF modules 'directly' through cable
« Reply #4 on: January 26, 2017, 03:27:19 am »
You can easily reduce the power coming back out of that splitter/combiner by adding a attenuator on it. In almost all cases they are not directional so it doesn't matter if its transmitting out or its getting power back trough it.

So put for example 30dB attenuators on each port of your splitter and the transmitting power gets cut down by 30dB as it goes in, then maybe the splitter looses 15dB and gets attenuated by 30dB again as it comes back out to the receiving device in total 75dB of loss. And that's something that is probably a reasonable signal strength seen in every day use.

This will work and I doubt a real splitter/divider is needed; the return loss through the attenuators will be so high that BNC-Ts will work.  Connect all of the devices through 30dB (40dB?) attenuators to a single length of coaxial cable like the old Ethernet 10Base-2 network.

Of course now thanks to the FCC you will need a reverse SMA adapter for every device.  That rule sure worked out well.

 

Offline martinmunkTopic starter

  • Newbie
  • Posts: 2
  • Country: dk
Re: Connecting RF modules 'directly' through cable
« Reply #5 on: January 27, 2017, 08:06:20 am »
Thanks for the input, people!

I am aware that i need to reduce the output power a lot in some way.

If nothing else could be done, i was thinking of mounting 10 antennas inside a big aluminum block. But I was afraid that that would do more damage than good.

This will work and I doubt a real splitter/divider is needed; the return loss through the attenuators will be so high that BNC-Ts will work.  Connect all of the devices through 30dB (40dB?) attenuators to a single length of coaxial cable like the old Ethernet 10Base-2 network.

Of course now thanks to the FCC you will need a reverse SMA adapter for every device.  That rule sure worked out well.

Berni and David:
This solution seems worth a try!
Do you know any good place online to shop for BNC adaptors and such. Preferrably in the EU. (We're in Denmark)

Was not aware that the pesky RP-SMA connectors was a FCC thing.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17498
  • Country: us
  • DavidH
Re: Connecting RF modules 'directly' through cable
« Reply #6 on: January 27, 2017, 02:05:11 pm »
This will work and I doubt a real splitter/divider is needed; the return loss through the attenuators will be so high that BNC-Ts will work.  Connect all of the devices through 30dB (40dB?) attenuators to a single length of coaxial cable like the old Ethernet 10Base-2 network.

Of course now thanks to the FCC you will need a reverse SMA adapter for every device.  That rule sure worked out well.

Berni and David:
This solution seems worth a try!
Do you know any good place online to shop for BNC adaptors and such. Preferrably in the EU. (We're in Denmark)

Offhand I do not.  I suggested BNC connectors because they will probably be the least expensive and most commonly available.

To save the cost of all of the extra T adapters and connectors, I would directly solder the ends of a bunch of coaxial pigtails to a copper clad board to make a resistive wye splitter.  If you find a source of inexpensive BNC patch cables, then they could be cut in two to make two coaxial pigtails each.

https://www.microwaves101.com/encyclopedias/resistive-power-splitters

Quote
Was not aware that the pesky RP-SMA connectors was a FCC thing.

The idea was that by requiring non-standard connectors, it would prevent users from combining transmitters and antennas to produce unapproved combinations.  It did not work and the practical result was twice as many RF connectors.
« Last Edit: January 27, 2017, 04:12:02 pm by David Hess »
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5057
  • Country: si
Re: Connecting RF modules 'directly' through cable
« Reply #7 on: January 27, 2017, 04:06:21 pm »

The idea was that by requiring non-standard connectors, it would prevent users from combining transmitters and antennas to produce unapproved combinations.  It did not work and the practical result was the twice as many RF connectors.

Oh so that's where that god damn RP SMA came from. I never understood why one would do such a thing to one of the most standard RF connectors out there.

So yeah same deal with security torx. It only works for a little bit but then they soon end up in every multibit screwdriver set and becomes just another screw type that you need to keep around a fitting screwdriver bit for.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17498
  • Country: us
  • DavidH
Re: Connecting RF modules 'directly' through cable
« Reply #8 on: January 27, 2017, 04:39:38 pm »

The idea was that by requiring non-standard connectors, it would prevent users from combining transmitters and antennas to produce unapproved combinations.  It did not work and the practical result was the twice as many RF connectors.

Oh so that's where that god damn RP SMA came from. I never understood why one would do such a thing to one of the most standard RF connectors out there.

So yeah same deal with security torx. It only works for a little bit but then they soon end up in every multibit screwdriver set and becomes just another screw type that you need to keep around a fitting screwdriver bit for.

This is also why most or all laptop models have an approved and very short white list of internal WiFi adapters that will work enforced by the BIOS.  The only approved adapters are the ones that the laptop was tested with for FCC certification.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf