Scheduled maintenance: Community will be offline Monday June 17th, 1:00 AM - 6:00 AM (PT)
New Member
Posts: 6
Registered: ‎11-03-2015

Sufficient Signal Strength

We have 2 Powerbeam M5 300 dishes operating a distance of 1.84km (1.15 miles). We believe it's fairly good line of site, although there might be some light foliage in the path.  Signal strength is around -74 dBm, and we have tweaked all the wireless settings to give maximum stability, albeit over just 4Mb/s or so. Ping times over the link are consistent at between 5ms and 15ms, with no loss.

 

My questions are:

 

1. Is this signal strength sufficient, or should we be looking to get it higher? I can't be certain whether it's the vertical that is out slightly (we don't have the ability to change this now, at one end at least), or the possibility that we might have foliage somewhere in the LoS.

 

2. Would a larger dish at the end we still have easy access to (AP, and the site that is sending most of the data) help?

 

(Main status display screenshot attached)

status.png
Senior Member
Posts: 3,505
Registered: ‎07-17-2010
Kudos: 839
Solutions: 201

Re: Sufficient Signal Strength

I'd say you have some obstructions.

 

You want a signal of -50 to -55 

 

You can increase gain and tx power, but obstructions need to be avoided first, if at all possible.

 

What country code is this radio set to?

New Member
Posts: 6
Registered: ‎11-03-2015

Re: Sufficient Signal Strength

Country code is GB (or UK?!), TX power is set to the max 14 dBm (although is this country specific?)

Senior Member
Posts: 3,505
Registered: ‎07-17-2010
Kudos: 839
Solutions: 201

Re: Sufficient Signal Strength

It is country specific. Leave it as is.

 

You need to elevate both, or at least one end to get a clear path.

 

That is the only way to get a solid link.

New Member
Posts: 6
Registered: ‎11-03-2015

Re: Sufficient Signal Strength

They've been elevated as much as possible, and it hasn't made much difference. 

 

The link is totally stable, it's only the signal strength (and constrained bandwidth, as a handoff to stability) that concerns me. So wondering if it actually is an issue, and (if it is) whether a bigger dish will help.

Emerging Member
Posts: 97
Registered: ‎10-08-2012
Kudos: 60
Solutions: 5

Re: Sufficient Signal Strength

The CCQ is good, which suggests the frequency does not have much interference.

The chains are 5dB apart, which suggests poor LOS.  When you refer to LOS, do you mean visual LOS or Radio LOS? Have you take into consideration the Fresnel Zone and ensured nothing is in the way of it to either side or below/above?  Are both antennas installed in the clear and away from any nearby obstructions off to one side or above/below?

Changing the radio at one end for one with a larger antenna will have some improvement, but it would need to be substantially larger to get that extra 20 or so dB you will need!

 

I would agree with what others have said, it looks to me that both ends need raising or moving into a better location if you wish to get more throughput.

 

Ron Touw - Ubiquiti Certified Trainer at LinITX
Get your Ubiquiti equipment and Training: LinITX.com
Official UK Ubiquiti Master Distributor
IRC channel: #ubnt on irc.z.je (IPv4) or 6.irc.z.je (IPv6)
New Member
Posts: 6
Registered: ‎11-03-2015

Re: Sufficient Signal Strength

Thanks, that's a great help. TBH the throughput isn't of concern. We need 2Mb/s, and we're getting over twice that.

 

So I guess I'm saying how important is signal strength really, given that we have the throughput we need, and it seems very stable (even throughout the heavy snow we have just had here)? Is there some crazy atmospheric condition that could lower the signal strength further, and cause problems due to us not having perfection in normal conditions?

Emerging Member
Posts: 97
Registered: ‎10-08-2012
Kudos: 60
Solutions: 5

Re: Sufficient Signal Strength

Greater signal will give you higher data rates, thus in turn higher throughput.  But it also gives you some margin for any drops outs due to atmospheric conditions.  Ice and snow does not usually cause any degradation at 5GHz. Melting ice on the surface of the radome however will. Heavy rain has actually been known to increase the signal level especially on long links. However as your link is very short, I doubt that anything in the atmosphere is going to make much difference to the signal levels.

 

What will cause degradation in the future is if the low signal strength is due to physical obstructions, such as trees in the Fresnel Zone.  Trees grow and therefore it will only get worse.  Unless you hire the services of someone with a chainsaw Man Happy

 

 

Ron Touw - Ubiquiti Certified Trainer at LinITX
Get your Ubiquiti equipment and Training: LinITX.com
Official UK Ubiquiti Master Distributor
IRC channel: #ubnt on irc.z.je (IPv4) or 6.irc.z.je (IPv6)
New Member
Posts: 6
Registered: ‎11-03-2015

Re: Sufficient Signal Strength

Update - running pingplotter over the link has shown that we get packet loss for about 30 seconds, every 5 minutes. Almost exactly. The signal strength stays the same.

 

Any idea what this could be? It's very regular!

Ubiquiti Employee
Posts: 11,657
Registered: ‎04-14-2017
Kudos: 2188
Solutions: 335

Re: Sufficient Signal Strength

What does your spectrum look like?
New Member
Posts: 6
Registered: ‎11-03-2015

Re: Sufficient Signal Strength

Expand?

Ubiquiti Employee
Posts: 11,657
Registered: ‎04-14-2017
Kudos: 2188
Solutions: 335

Re: Sufficient Signal Strength

Are you able to see via spectrum analysis on the radios whether anyone else is operating on the same channel as you?
Emerging Member
Posts: 97
Registered: ‎10-08-2012
Kudos: 60
Solutions: 5

Re: Sufficient Signal Strength

Yes, probably poor Line of Sight. The chains are 5dB apart.  Possibly trees moving in the wind, rain water on leaves can make signals move up and down, but it all depends on what is in the Fresnel Zone.  Only you know that.

Ron Touw - Ubiquiti Certified Trainer at LinITX
Get your Ubiquiti equipment and Training: LinITX.com
Official UK Ubiquiti Master Distributor
IRC channel: #ubnt on irc.z.je (IPv4) or 6.irc.z.je (IPv6)