1

For a larger project where I basically want to boost phone signals indoors using an external antenna I have devised a small experiment to make sure I get the basics right before I expend a lot of energy.

Basically, I have made an enclosure to reduce signal strength in which I put a cell phone displaying the current RSSI or dBm and a small-scale version of the external and internal antenna setup. Basically just two small antennas connected by a wire.

The idea was that I should be able to measure a difference in signal strength inside of the box when putting a small antenna connected to another antenna outside of the box. This would then validate the feasability of using the bigger antennas.

The setup

Step 1 was to make an enclosure that would block (some) signals. A card board box wrapped in aluminium foil was no true Faraday cage, but when closing the lid it managed to reduce the signal strength by 15dB, which is noticable.

Signal reducing enclosure works Enclosure to reduce signal strength

Step 2 was then to verify that these antennas would boost the signal inside of the box. This is where I fail, somehow. While closing the lid of the box made the signal go from -90 to -105 on the dBm scale, putting the antenna in there made absolutely no difference - no matter how long I waited.

I suddenly realized that I might be reading the wrong figures, as I still had 4G enabled (the antennas are both for about 900 MHz, while 4G is 1800 and 2100 MHz in Norway), but disabling 4G and strictly running on 2G made no real difference.

So this leaves me wondering: what am I doing wrong? Are there some basic assumptions I have made on how this should work that are just plain wrong?

Debugging steps:

  • I have verified that there's no short in any of the links from the black whip up until the N-connector using a voltmeter (0.6 V reading).
  • I have turned off 4G
  • I have verified that I lose 12-15dB of signal when closing the lid
  • Tech specs on the long whip (900 MHz range)
  • Tech specs on the stick antenna (900 MHz range)
  • The splitter box used to connect the two male cable connectors seems to be fine too when checking using a voltmeter.
oligofren
  • 321
  • 3
  • 9
  • A passive re-radiator requires a very short cable. Attenuation in the cable can be substantial. – user57037 Jan 29 '18 at 23:53
  • I understand that, but we are talking about a mere 3m of cable. Even with a flimsy cable like pictured, that's usually not more than handful of dB at such a short length. Surely it should show some improvement in the signal reading? – oligofren Jan 29 '18 at 23:59
  • I mean, compared to a 15dB drop there should be something to show the signal has improved by an ever so small amount, right? – oligofren Jan 30 '18 at 00:06
  • It looks like you are embarking on a device called "Cellular Signal Booster", e.g. https://www.signalboosters.com/ These devices are not so simple to design, and they are priced at $250 - $1000 and more. I am afraid there is a reason behind that. – Ale..chenski Jan 30 '18 at 04:31
  • 1
    @AliChen, that's incorrect. Those are active devices. This concerns passive amplification. Basically you would put one of those where my signal splitter is. Problem is that active GSM amplification is illegal in Norway for anyone that's not a phone company, as it will disturb reception for everyone else. Thus passive amplification using a high gain antenna is usually your only bet. – oligofren Jan 30 '18 at 05:16
  • Why do you have a splitter box? That could be causing lots of trouble. I think you need to try this with very short cable and no splitter. To be honest, I think you should just try it full scale and quit messing around. There are a lot of potential issues with the miniature setup that prevent it from being predictive of results with the full-scale setup. For one thing, in the full scale setup you will use a directional antenna outside, so there is that. Also, the dimensions of the box could possibly be causing strange resonances inside. – user57037 Jan 30 '18 at 05:21
  • The splitter is there as I will be connecting a remote control of my heating directly, as well as an internal antenna. I tried doing this "large scale" in my remote cabin once, and couldn't see any difference in signal reading, so I wanted to see if I could make a controlled experiment at home where I more easily could control the variables before embarking on it again. I was also starting to wonder if the high gain antenna worked at all, since no change in signal could be detected. The answer mentioning directionality on the internal antenna might be an explanation. – oligofren Jan 30 '18 at 05:35
  • 1
    The splitter is for sure going to cost you several dB. Probably over 3.5dB. Think about it. It splits power. Also, what is the impedance seen looking in from the "big" antenna side when one of the splitter ports is open circuit? So the unterminated splitter could possibly be creating a mismatch. To give this a chance, I would definitely try to find a very short cable and do a direct connection with no splitter. – user57037 Jan 31 '18 at 01:55
  • 1
    @mkeith That's a good point. Thanks for that! For this experiment I just needed the splitter to connect the two male SMA connectors from each antenna. I'll see if I can find a connector with female in both ends. – oligofren Jan 31 '18 at 12:27

1 Answers1

1

The large antenna is collecting a portion of the cell tower's energy which falls within its "antenna aperture". Your system is then radiating this energy from the small antenna in all directions.

(Not really, of course, but you see the point).

Only a small amount of this power is being re-radiated in the direction of the cell phone's antenna.

I expect you could detect the improvement in signal strength if your chamber had a higher level of attenuation. But as it is it's getting swamped out by the direct signal which is coming from the cell tower.

bitsmack
  • 16,931
  • 9
  • 54
  • 111
  • Thanks, that makes sense. Given that the external antenna receives 1W of power, only 1W/x, where x is some integer, will reach the cell phone. Making x smaller by directing the energy in one direction would then probably give better results. Too bad smartphones can't be directly connected by cable: that would make x closer to 1, I assume. – oligofren Jan 30 '18 at 05:27