3

tl;dr For minimal signal loss: a longer cable is better than solid obstructions. Right?

I would like to enhance (the non-existing) phone signal indoors and have been thinking of connecting an external antenna to one inside. Basically:

cell phone <-> whip antenna <-> RG-58 cable <-> aerial

So I was thinking where to place the indoors antenna. Should I try to minimize cable length or obstructions? I might have found a conclusion by researching for this question, but I would still like to verify my thinking.

As RG-58 has about a loss of 0.3 dB/m at 960 MHz (GSM), and a 15m cable run would mean about 4.8 db loss according to this online calculator, or about 1/3 (1W in gives 0.3W).

On the other hand, solid wood dampens the signal by 5-12 db. That means even a 2m stretch of cable to minimize loss would be for nothing if I had an additional timber frame or door for the signal to pass through.

Is my thinking correct?

P.S. Follow-up question on using 75 ohm vs 50 ohm cable here

oligofren
  • 321
  • 3
  • 9

1 Answers1

1

Your observation seems correct but choice of cable is poor. Except doors are dry wood and that makes a big difference.

Use SAT or CATV coax with satellite dish but LNA removed but located above chimney.

You need a high gain direction rooftop antenna in order to act as a passive repeater to be much stronger than the direct phone signal levels, otherwise, it will be worse by adding another multipath fading error.

Old cheap sat. DTV dishes were only 5dB gain. Bigger ones , more precise ($) could get more dB gain. But this shows the ideal diversity ( beamwidth) vs antenna gain tradeoff. A whip is almost isotropic but not down the end of the conductor, so close to 1dB gain. It assumes direct line of site, so the miles is theoretical.

enter image description here

schematic

simulate this circuit – Schematic created using CircuitLab

This is like an ascii drawing of a house, ideal antenna aimed at tower above trees, with moot point about 75 vs 50 ohms considering coupling impedance between phone and whip antenna. SO I suggest you will need to wrap a few turns of spring telephone coil around mobile phone to couple energy to phone patch antenna.

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 185
  • 1
    RG-58 (or RG-223) is the standard choice for most telecom stuff. Isn't satellite cables also of wrong impedance? It should be 50 ohms, not 75 ohms, supposedly. I might also attach a directly connected device, as well, using a splitter and then I guess it would matter. One usually gets ready made RG-58 with SMA connectors online. – oligofren Dec 19 '17 at 17:17
  • Both terrible choices above with RG223 coax cable has an attenuation at 1 GHz of 13.4dB/100'. As I said go for cable above with attenuation of 3dB max including , connections , & return loss or else what's the point? Your distance and antenna gain, impedance specs are what? d= to go from above roof to main floor direct path? – Tony Stewart EE75 Dec 19 '17 at 18:54
  • Coupling impedance between whip and phone; not sure I follow? I am not physically connecting an iPhone to the whip (so where's the coupling?). If I were to connect anything it's a GPRS shield to remote control heating. Those have standardized SMA-connectors. Not sure if I can get one for any old cable? – oligofren Dec 20 '17 at 01:52
  • I have looked up attenuation tables, so I see your point on RG58, as it can have an attenuation of everything from 14 (Belden 8240) to 23 dB (Belden 8259) per 100', whereas Belden's RG6A has about 10dB loss per 100' (Wikipedia says 6dB/100' but haven't found any real cable matching that). RG6 is still a 75ohm cable, with the problems of standing waves when connecting to 50ohm equipment, but I'll ask a seperate question on that to not mess this one up. – oligofren Dec 20 '17 at 12:37
  • 1
    Tony's schematic shows a coil of wire wrapped around a phone, which would capacitively couple the signal to your phone's antenna with some impedance. He did it that way because most phones don't have external antenna connectors. GPRS shields do have antenna connectors, but weren't mentioned in the original question. – remcycles Dec 20 '17 at 15:07
  • 1
    @oligofren higher impedance coax will be closer to the impedance of free space between phone and your whip antenna, which is why I suggested lower impedance coupling by mutual BH field coupling. Crude but much better than Friis Loss. The OTA was intended for lightning rod. – Tony Stewart EE75 Dec 20 '17 at 16:21
  • I agree in hindsight I could have added the restraint that this was also supposed to be used with a directly connected (via a splitter) GPRS shield, but I was thinking it would have complicated the question needlessly. Seems I should have, given the response, so sorry about that. I wasn't expecting such creative/knowledgeable solutions :-) Also, I could have added that the phones should work wirelessly/not directly connected, but I thought that would be implicit due to the antenna. I am afraid coiled wire around the phones would put a lot of my guests off, but it's certainly clever. – oligofren Dec 20 '17 at 17:08
  • Btw, I was really wondering about all this 75 vs 50 ohm thing, so I made another question that got a lot of good answers. For any latecomers that might be in my shoes it might well be worth checking out: https://electronics.stackexchange.com/a/345977/93601 – oligofren Dec 20 '17 at 17:10