dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
560
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

Ducting or...?

I have a Nanobridge PTP link over water. 140' on the AP side 292' on the CPE side. Some nights the signal will drop from -59 to around -78 to -80. Link is 5.3 miles 2.5 of which is over water. If I change bands so go from 5800 to 5160 while this is occurring it will jump from -78 to -80 down to -65 and be happy until the next morning when it returns to -59. This happens only at night and well after sun down - 11pm ish.

John Galt6
Forward, March
Premium Member
join:2004-09-30
Happy Camp

John Galt6

Premium Member

You're seeing an inversion layer develop rather than traditional ducting. The effect is about the same, though.

What are the daily day/night temps like?
nosomo
join:2008-12-17

nosomo to wirelessdog

Member

to wirelessdog
Why not use airselect -- give it a channel list -- increase the delay before switching channels -- It should continually use the better channel no matter which it is.
prairiesky
join:2008-12-08
canada

prairiesky to wirelessdog

Member

to wirelessdog
I get that every fall for a few days. PITA. Inversions.

TomS_
Git-r-done
MVM
join:2002-07-19
London, UK

TomS_ to John Galt6

MVM

to John Galt6
Yeah we had a bad link that exhibited this, particularly during the warmer months. It ran down the middle of a peninsula, but only a couple of km from the ocean. As night drew, and particularly towards midnight, we'd be getting alerts left right and centre due to an inversion layer.

Such to the point that if I was on-call I would just turn off alerting because there was stuff all I could do about it, I might as well get proper sleep and be woken by actual network outages.

We eventually upgraded to a space diversity link which improved the situation massively.

Later on in my time at that company I started to write some smarts in to the monitoring system. Alarms would be buffered between midnight and 6am, so if there had been events during the night, the first time you'd hear about it would be 6am, and by that stage most likely you'd have had a decent enough night sleep to want to care about it, and would still have a couple of hours to head out to a site to make repairs before the day kicked off proper.
raytaylor
join:2009-07-28

raytaylor to nosomo

Member

to nosomo
said by nosomo:

Why not use airselect -- give it a channel list -- increase the delay before switching channels -- It should continually use the better channel no matter which it is.

That is a good idea
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

Last I used airselect it wreaked havoc for my gamers. It would drop their connections. Has it improved?
OHSrob
join:2011-06-08

1 edit

OHSrob

Member

said by wirelessdog:

Last I used airselect it wreaked havoc for my gamers. It would drop their connections. Has it improved?

Its only a suitable choice in my testing for if you only have very light users.

edit: The WS units seem to be able to change channel's faster.
nosomo
join:2008-12-17

nosomo to wirelessdog

Member

to wirelessdog
Being that this is a ptp link, amount of users on it shouldn't matter. Default change is 3 seconds. It'd be nice if we could feed airselect variables specifically for this instance where a 'default' channel fades beyond a certain signal and airselect shouldn't try it again for say 30 minutes. In that instance keeping a very short channel change would allow the radio to change, check signal and immediately change back with minimal impact to the link.

I have a very noisy 2.4 environment and airselect is the only way I can get a stable link. It's constantly switching to channels with minimal broadcast and keeps an otherwise unusable link, usable. I've had no complaints for those on the link and the PPPoEs stay live.

DaDawgs
Premium Member
join:2010-08-02
Deltaville, VA

DaDawgs to wirelessdog

Premium Member

to wirelessdog
Hi WDog;

With a link that has one end 160 feet higher than the other end, especially over water, you are going to see this problem more often than with one where both ends are closer to the same height.

This is because the inversion layer rises and falls during the daily cycle. Other factors, most notably blustery winds, also affect whether or not this inversion layer even forms. It is probably not news to anyone here that inversion layers tend to form when the air is calm and there is a source of heat (water) under a cold clear sky.

What I think is happening from your description is that the inversion layer is moving to an altitude between the height of your high node and the height of your low node. Then the inversion is acting as a reflector and holding the signal below it in the warm, moist duct formed between it and the water surface.

I think you would do well to move the far end to a lower point on your tower, if that is possible.
wirelessdog
join:2008-07-15
Queen Anne, MD

wirelessdog

Member

I thought that shots over water should always be at different heights???

DaDawgs
Premium Member
join:2010-08-02
Deltaville, VA

DaDawgs

Premium Member

said by wirelessdog:

I thought that shots over water should always be at different heights???

I don't know why that would be true. I hope that I am not the person who suggested that. =) I might well be but if I am, I was wrong.

In my thinking, in this case, it is the different heights that are causing the issue. One end is more than twice as high as the other end. That is a big difference. Easiest way to check it to install one at 140 to 160 on the high end and run them in parallel for a month during the season you see the most problems.

=)

You have seen mirage on the highway I expect. The same thing happens with radio waves. Mirage is caused by a temperature inversion which reflects light. Radio experiences the same thing.

So suppose that one end is under the mirage and the other is above it. Most of the signal is reflected away from the other end. You loose twenty or thirty dB.

billaustin
they call me Mr. Bill
MVM
join:2001-10-13
North Las Vegas, NV

billaustin to wirelessdog

MVM

to wirelessdog
said by wirelessdog:

I thought that shots over water should always be at different heights???

If I remember correctly, shots over water are supposed to work better with horizontal polarity.

Atmospheric ducting is a phenomenon that essentially creates a tunnel in the sky that will carry a signal much farther than it was ever intended to travel.
LLigetfa
join:2006-05-15
Fort Frances, ON

2 recommendations

LLigetfa to DaDawgs

Member

to DaDawgs
said by DaDawgs:

said by wirelessdog:

I thought that shots over water should always be at different heights???

I don't know why that would be true.

WRT shooting over water, when the beam is parallel to the water, then there can be a destructive reflection off of the water. When there is a thermal inversion over either flat land or water, there can also be a destructive reflection off of the "mirage". These destructive reflections are sensitive to the height of the beam and also sensitive to frequency. The size of the fresnel zone changes with frequency so the odd/even additive/destructive reflections change with it. That likely explains why changing from 5800 to 5160 helped.

In this particular case there is significant height difference between ends, but there can be reflections at any angle. Odd/even, additive/destructive fresnel issues can often be mitigated by varying the height to move off of a destructive fresnel but that only works if the reflective surface is not changing. Ocean tides and thermal inversion create a moving reflective surface. Space Diversity is sometimes used to mitigate that. Circular Polarity will also reduce destructive reflections. Diagonal Polarity is one more thing to try.

DaDawgs
Premium Member
join:2010-08-02
Deltaville, VA

DaDawgs

Premium Member

Yeah, I forgot about tidal effects... DERP. ; I am probably the one that told him different heights, I remember going on about it with some sites that had fade problems twice a day that matched the tide change.

You nailed it here. Well answered.