Hello everyone,
I'm having a big interference problem on two of my APs.

There are A LOT of competitors and every channel except two is used and I can see them both client and AP side with signal > -75

My competitors are near me with their AP.

I'm using two free frequencies for two sectors, on a RB433 with two wireless card.
Protocol is nv2 and band 5GHz-N, 20MHz channel.

Some clients are getting high CCQ (>90%) and good speed (15-18mbps), others terrible CCQ (<40%) and low speed (< 2Mbps).

I scanned on the CPE side and I'm the only one who is using these frequencies.

I tried with nstreme instead of nv2, better speed but disconnections. 802.11 low speed.

Can you please tell me how do you solve this kind of problems?
I'm sure everyone had to deal with RF interference

Thank you!

Fortunately, the FCC just authorized the expanded use of the UNII 1 band so you can change frequencies to that band if your clients will support it. Changing freqs is the only real solution. If the air space is saturated all you are going to do is beat your head against the wall.

Thank you for your reply!
Unfortunately also UNII 1 band channels are used, and i already tried to beat my head against the wall, but it didn't solve the problem

i already tried to beat my head against the wall, but it didn't solve the problem

I've been there too

Unfortunately, without clean air space everything becomes a trade off. You can decrease the channel size. That will help considerably, but the trade off is throughput. Have you ever used a wifi router in a crowed apartment complex? There may be dozens of routers, maybe even a hundreds of them, but for the most part, your stuff will work with your router. That's because of the distance and and the power. If every bit of spectrum is dirty, you are going to have to increase the signal to noise ratio for everyone of your clients and your APs. You can accomplish this by:
Reducing the channel size
Using more directivity with your antennas
Installing customers closer to your APs
Using increased power

Many people will disagree with this next statement, but spend some time thinking about it. Broadcasting from tall towers and from the top of tall buildings is the worst way to build a network. An experienced network engineer will only used the high space for BHs. Use the landscape to your advantage. When you can drop down below the tree line everybody else's signal is attenuated. That's how you deliver good service for a long period of time. Make your hopes small, always use as much directivity as possible, and use the landscape to your advantage.

Changing the antenna polarity in crowded locations has helped me in the past.

...........................................................

Many people will disagree with this next statement, but spend some time thinking about it. Broadcasting from tall towers and from the top of tall buildings is the worst way to build a network. An experienced network engineer will only used the high space for BHs. Use the landscape to your advantage. When you can drop down below the tree line everybody else's signal is attenuated. That's how you deliver good service for a long period of time. Make your hopes small, always use as much directivity as possible, and use the landscape to your advantage.

Totally agree, once you have calculated height to clear freznel zone then look around and use a disadvantage to your advantage by placing say on the side of a building facing your clients rather than on top which would pick up unwanted signals also.

Change polarity is another good method to use

You can use closer sector antennas (40-60degrees) - Integrated with rb912. For example we use antennas with AL box
http://www.raytra.com/en/5g-anteny-parabola.php

433 - discard

You can use closer sector antennas (40-60degrees) - Integrated with rb912. For example we use antennas with AL box
http://www.raytra.com/en/5g-anteny-parabola.php

433 - discard

What is wrong with 433?

I'm using two free frequencies for two sectors, on a RB433 with two wireless card.

This is problem in noisy enviroment!
Better is using --> One antenna with one routerboard

I'm using two free frequencies for two sectors, on a RB433 with two wireless card.

This is problem in noisy enviroment!
Better is using --> One antenna with one routerboard

+1

Thank you all guys, great feedback about your experiences!

Changing the antenna polarity in crowded locations has helped me in the past.

...
Change polarity is another good method to use

What about using MIMO antennas?

Thank you all guys, great feedback about your experiences!

Changing the antenna polarity in crowded locations has helped me in the past.

...
Change polarity is another good method to use

What about using MIMO antennas?

If your interference is on say vertical polarity then there is no point using Vertical?

MIMO use both vertical and horizontal, is it a problem?

MIMO use both vertical and horizontal, is it a problem?

Some don't use V+h but slant polarity but first establish which polarity gives you the least amount of interference and plan from there.

The dual polarity only come into to play with integrated antennas. For exanmple, you could use two horizontal antennas instead of a dual polarity antenna and it will work just fine.

http://www.rfelements.com/assets/sheets ... eetEN2.pdf

If I use this and on client enable both or only horizontal polarities, does it work like using only horizontal?
I would like to use it as I have lots of clients already set on vertical polarity only!

You mentioned

I'm having a big interference problem on two of my APs.

are they the sectors attached to 433,

Yes, they are attached to 433, i meant "two of my sectors on an AP"

Yes, they are attached to 433, i meant "two of my sectors on an AP"

Then the the first step in reducing interference caused by two radio cards attached to a 433 is to as others will advise is use 1 routerboard + 1 radiocard to a sector antenna

The problem is this place is very uncomfortable and I would like to reduce the number of RB to the minimum.
Do you know any trick to shield them better?
Is using 1RB => 1 Radio the only real solution?

Why is MikroTik selling routerboards with 2, 3, 4, 5 radio cards if using these slots is an issue?

The answer to that is that its not always an issue. In a clean air environment, with the wireless cards spread out in frequency or even spread across multiple bands, the performance is very good.

The problem is this place is very uncomfortable and I would like to reduce the number of RB to the minimum.
Do you know any trick to shield them better?
Is using 1RB => 1 Radio the only real solution?

Why is MikroTik selling routerboards with 2, 3, 4, 5 radio cards if using these slots is an issue?

A very good question often asked, lets use 433 as a example, it could use three radio cards 900/2.4/5.8 but not 2.4/2.4/2.4 or 5.8/5.8/5.8.....

There is no good trick to shielding them but a much better and more efficient method of maximizing TX radiated RF power is to use integrated sectors that have a radio compartment to place the Mikrotik board.

I have a shielded container for RB and i put it 1m away from the sectors, attached to it with two shielded coaxial cables.

So, you suggest:
1 RB => 1 Radio -> When there are competitors and there isn't clean air space
1RB => More radios => When I'm alone and there is clean air space
?

Exactly where is the noise located? I mean, the problem is the radio card, the pigtail, the sector...?

I have a shielded container for RB and i put it 1m away from the sectors, attached to it with two shielded coaxial cables.

So, you suggest:
1 RB => 1 Radio -> When there are competitors and there isn't clean air space
1RB => More radios => When I'm alone and there is clean air space
?

Exactly where is the noise located? I mean, the problem is the radio card, the pigtail, the sector...?

No - It's just good installation practice for performance to use 1 RB=1 radio - for all locations and regardless of clean air space or not - many other users will advise the same, now using integrated AP or not it's you choice but you will have more transmitted power when using integrated and not using the 1M RF loss loss LMR ,etc. cable, quoted cable loss and actual measured on site insertion loss when using RF connection lead can be in dB's and not say 0.15dB as in the data sheet,

Noise pickup is due to the proximity effect of the radio cards antenna terminals and PCB tracks, which cannot be screened 100%, when one radio card is in transmit mode the other maybe in receive mode, internal RF signal pickup is going to happen...............

Thank you for your reply, I'm learning a lot
What product do you suggest for integrate AP?

Thank you for your reply, I'm learning a lot
What product do you suggest for integrate AP?

I have a preference for MIMO AP's
http://www.itelite.net/en/Katalog/MIMO- ... HVBOX.html

Others may advise of equally good AP's

P.S. Don't forget to give Karma for good advice given on the forum

its possible to use X polarity antennas in mimo

Go to rfarmor.com

They work!

With multiple cards in one board keep your tx power at 20 or below. Helps tremendously.

With multiple cards in one board keep your tx power at 20 or below. Helps tremendously.

Multiple cards using the same frequency band in the one board is a bad idea.

No argument, but once the cards are there, you can swap the AP to get it working or you can drop the TX power.

I have 3 R52H in a 433 since I started in 2008. The self interference issue has been better or worse depending on the version of RouterOS, but since I had several pieces of hardware in the air with multiple cards in single boards I learned what to do to make them work.

Essentially, if you scan from one card and see the other in the same band with a signal greater than -40dB you will have issues. For 2GHz and 5GHz, setting the tx power below 20dBm gets the signal level close to -40 depending on the cards, pigtails, etc.

I've found that low power cards have much less self interference issues than high power cards. R52 work fairly well in the same board, R52H do not. I haven't tried this with newer cards without diversity chips. The lack of a diversity chip should help shield the signal a bit more inside the case.

I've also learned that running multiple XR5 in a single board has much less of a problem than the R52H. I'm not sure if it is due to the pigtail/coax (MMCX vs UFL) or the lack of a diversity chip. Either way, at default tx power, two XR5 in a single case see each other around -40dB in the 5GHz band. Two R52H see each other at -20dB, I've even seen signal in the teens before.

No argument, but once the cards are there, you can swap the AP to get it working or you can drop the TX power.

I have 3 R52H in a 433 since I started in 2008. The self interference issue has been better or worse depending on the version of RouterOS, but since I had several pieces of hardware in the air with multiple cards in single boards I learned what to do to make them work.

Essentially, if you scan from one card and see the other in the same band with a signal greater than -40dB you will have issues. For 2GHz and 5GHz, setting the tx power below 20dBm gets the signal level close to -40 depending on the cards, pigtails, etc.

I've found that low power cards have much less self interference issues than high power cards. R52 work fairly well in the same board, R52H do not. I haven't tried this with newer cards without diversity chips. The lack of a diversity chip should help shield the signal a bit more inside the case.

I've also learned that running multiple XR5 in a single board has much less of a problem than the R52H. I'm not sure if it is due to the pigtail/coax (MMCX vs UFL) or the lack of a diversity chip. Either way, at default tx power, two XR5 in a single case see each other around -40dB in the 5GHz band. Two R52H see each other at -20dB, I've even seen signal in the teens before.

Interesting about the XR5's have you tried with one XR5 if using say 5520 to set frequency scan to 5500-5540
and another XR5 if using say 5700 set frequency scan 5680-5720, then set TX to default and check if interference effect is reduced?

Hi everybody
I have a ptp link in very noisy environment . Tx rx are -49/-50 but throughput is 1Mbit .
Is there any solution ? ( I know I should change frequency . Not this solution )

Hi everybody
I have a ptp link in very noisy environment . Tx rx are -49/-50 but throughput is 1Mbit .
Is there any solution ? ( I know I should change frequency . Not this solution )

Hi everybody
I have a ptp link in very noisy environment . Tx rx are -49/-50 but throughput is 1Mbit .
Is there any solution ? ( I know I should change frequency . Not this solution )

That depends what is real reason of your low throughput. You didn't share much info.
- what noise floor and SNR do you see on both ends?
- what CCQ is on both ends?
- as it is ptp link, what antennas do you use?

Distance?
Frequency?
Recommend speed