Small project with problems

[moa5505 1]

Hello everyone !

I am kind of new in this forum, this is my first post, but I have been lurking a lot these last few days.

I wanted to describe you a small project of mine and some problems I have been having. Maybe my project could give some ideas to others and/or maybe you could have an idea to solve the problems I currently have.

 

So I am currently living in a small apartment in a medium density residential area (some small building and individual houses around). When I arrived here I noticed there was a public access wifi (in which you had to enter some identification), since it was working quite well I didn't bother to pay for an internet connection. The first problem was that the signal was quite low (depending on where I was in my apartment I couldn't connect) and I had to enter a password every 2 hours...

I solved this by placing a raspberry pi in a specific corner in my apartment and programming it to check the connection every 10 seconds and automatically send my credentials to connect if no internet was available. This worked very well for a few months, but a few weeks ago the wifi I was using for this went offline... To solve this I bought a better receptor (Alfa AWUS036NH) and a better antenna (TP-link ANT2409A), boosted the transmission to 30dBm (1W) and pointed the antenna to another building. And it worked ! I had internet again !

But I noticed the speed of the connection was much slower than before it's usable, but it's very very slow ...
I can barely watch a 360p video without lag ...

 

So now that I presented my project I am exposing my problem.

I have a quite good signal (-70dBm), and since my transmission is made with 1W, the router should not have any problem in receiving my packets, but I have a very high rate of TX retries (300%). I tried to move the antenna a bit, but nothing seems to change this. I also tried to block the transmission rate at a lower bitrate (5.5MBit) hoping that it would solve my problem, but again no change ...

Does someone has an idea on why my setup is not working correctly ?

 

(I attached a screenshot of the main information of the connection)

[Admin 2545]

The antenna (TP-link ANT2409A),has only 6 dBi gain in directivity, try a stronger antenna made by you .. !!!

[moa5505 1]

3 minutes ago, Admin said:

The antenna (TP-link ANT2409A),has only 6 dBi gain in directivity, try a stronger antenna made by you .. !!!

The specification says it's 9dB, but I don't know about real case usage. Anyway, I'm not convinced that a better antenna would be a solution since I already have a pretty good signal (I don't know about the router though ...). I was thinking this is a problem caused by the distance (packets have to travel to much time and this delay causes the devices to think the packets were lost).

Why do you believe a better antenna could be a solution ?

 

[Admin 2545]

,,, it can be 9 dBi...!!!

,,,but try to rotate the antenna 45 degrees horizontally and change the position in different places...!!!

[moa5505 1]

18 minutes ago, Admin said:

,,, it can be 9 dBi...!!!

,,,but try to rotate the antenna 45 degrees horizontally and change the position in different places...!!!

Rotating the antenna kind of works, I get a lesser signal (-74dBm), but a much lower retry rate (120%) and watching a 360p video is now clearly possible. I will keep it like that for the moment.

Building an antenna sounds fun, but I don't have any tools at the moment. I will save your schematics though, thanks ! What are the specs of the one on top ? Is it a directional or omnidirectional antenna ?

[Admin 2545]

,,, you can try this with ALFA  AWUS036NH....

[swarg_eu 43]

the important thing isn't RSSI but SNR 

by rotating the antenna for 45 deg, you RSSI dropped bay 3dB but you SNR increased 

 

you seem to be a victim of CSMA/CD & CSMA/CA 

there are maximum 14 (Japan2) channels on 2,45 GHz WIFI and there is someone on the same channel 

Get a directional antenna ~18 dBi, with some luck the interfering signal is not in the direct line between you and the AP.

 

If needed with the 18dBi antenna, you have some space to get rid of the negative effects of the CSMA

by putting a 5~10 m long cable to attenuate the interference to a level where the CSMA won't react (it won't stop the transmitting, package drop)

[clanon 246]

if you don't get closser to -60dbm or even under -60... AND link QUALITY (PACKETS) OVER or CLOSE to 70% you won't have SPEED or stability

If you can  , try a reflector (ANY METAL) behind your antenna and the signal direction...

Edited September 27, 2020 by clanon

[moa5505 1]

8 hours ago, swarg_eu said:

the important thing isn't RSSI but SNR 

by rotating the antenna for 45 deg, you RSSI dropped bay 3dB but you SNR increased 

 

you seem to be a victim of CSMA/CD & CSMA/CA 

there are maximum 14 (Japan2) channels on 2,45 GHz WIFI and there is someone on the same channel 

Get a directional antenna ~18 dBi, with some luck the interfering signal is not in the direct line between you and the AP.

 

If needed with the 18dBi antenna, you have some space to get rid of the negative effects of the CSMA

by putting a 5~10 m long cable to attenuate the interference to a level where the CSMA won't react (it won't stop the transmitting, package drop)

Nice Thanks for the explanation !

I don't plan on changing my antenna (or buying cables) at the moment, but it could be a option. I'll keep that in mind.

[moa5505 1]

5 hours ago, clanon said:

if you don't get closser to -60dbm or even under -60... AND link QUALITY (PACKETS) OVER or CLOSE to 70% you won't have SPEED or stability

If you can  , try a reflector (ANY METAL) behind your antenna and the signal direction...

I will try to put some aluminum behind the antenna to see if it improves something, Thanks !

[clanon 246]

12 hours ago, moa5505 said:

I will try to put some aluminum behind the antenna to see if it improves something, Thanks !

DISTANCE and AREA/SHAPE (curved  angled) is IMPORTANT ...fiddle with it use a fast and sensible indicator of level quality 

[moa5505 1]

2 hours ago, clanon said:

DISTANCE and AREA/SHAPE (curved  angled) is IMPORTANT ...fiddle with it use a fast and sensible indicator of level quality 

You guys make it seems like some kind of art/witchcraft rather than an exact science 

[clanon 246]

5 hours ago, moa5505 said:

You guys make it seems like some kind of art/witchcraft rather than an exact science 

Electromagnetic Fields at HIGH Frecuencies...behave like  witchcraft ...(before simulations was hard to figure it out...and still is)