Go to Google Maps, and pick the exact locations of the antennas (assuming your’e in a high resolution maps area) and use the ‘measure distance’ feature to draw a line between the antennas. That line is your friend. Now, zoom on at each endpoint, and look for an easy-to-recognize landmark… a particular tree… a power pole, a neighbors shed, the corner of a house you’ll be shooting over… somethings that you can visually see from each endpoint. Then - aim each antenna, at the nearby visual ‘marker’ that you’ve picked out.
Here is an example – these two locations are 800 meters apart, but there’s enough clutter, and bushes, and trees here and there, that we can’t see each endpoint from the other. SO - I draw my line, and I zoom in, and bingo… from Alex’s house on the southwest end… I see that from that exact spot on his roof, I should be able to see the power pedestal across the road about 150 meters away. I also note that the neighbors lane, where it goes through her tree row, also happens to be inline. So, when I’m on Alex’s roof - I can’t see John’s house… but I can see the power pedestal, and I can see Lisa’s lane. I aim at those spots, and I’m very nearly perfectly aimed at John’s house too.
The number of times we’ve aimed over a granary in a field, or aimed ‘half way between those two power poles’ because that’s what lined up with our tower (which we couldn’t directly see) and at least then had a starting point to tweak from there.
So - maybe you can do something similar… get a good overhead map image, and look for something recognizable on the path which you can see from at least one end, and then you’ll know your reasonably close from there, at least as a starting point, on at least one end.
PS: My gut feeling is that you’re shooting into the dirt… not quite high enough to clear that bump - and that more height is going to be the winner. But that’s just a WAG.
Yes - a bridge is basically like a ‘long cable’. Just that this cable can have an IP address(es) which are used to log in and configurable options. But, if you give the radios 192.168.x.x or 10.x.x.x or whatever numbers you want… that’s fine. Think of the link just like a long Ethernet cable.
“My neighbor has a lift”. That is like finding a gold nugget. Get in the lift and go as high as you SAFELY can.
If the link connects then work your height down to where you can install an antenna.
If at that point the link does not work then you know it is time to give up and look for a different solution.
Don’t overthink things. Trust the math. Test the theory. Do the work.
Have you considered Craigslist or spending a day driving around looking for someone wanting to get rid of an old TV antenna tower. DIY Top Rail Mast, Flagpole, DIY Tilt Over Mast, etc.
Wow, thats a ground burner! You will definitely need some height on the low end of that link! Yes the first fresnel zone is 20ft across, but the third is close to 100ft across! So yes you can make a link with 20ft clearance on short links, anything where you are going over 2000ft will need more height and as pointed out, when dealing with trees 40ft away, you need to get above the forest floor.
Those ubnt antennas should be considered LOS only, get some KP dp yagis and use the slant 45 orientation, you will have better propagation through the trees.
You have been given everything you need in the above postes to make this work and then some!
Interesting about the “third” fresnel zone - why doesn’t Link Planner show this or stress it? The link is about 1/3mi or 2,000 feet (sorry about the units error).
I’m going to test the link part way with the battery powered antenna as LostInTheTrees recommends. See where it fails and then plan what to do about it.
Very bummed that the vendor sold me the UBNT Yagi when most people think they suck. LOS 900mhz antennas? Damn it.
I didnt say the ubnt yagis are LOS antennas but I did say that you should consider them as such as they are really poor performers compaired to what is available for similar and even lower price points.
Going into the woods to test is a good idea but it will only yield partial results as you are not fixing the link issue but rather making a different link altogether.
Understanding Fresnel zones is just part of our job as WISPs, knowing this helps us figure out what we did wrong and link planner does not calculate the second or third zones as for a LOS link they dont usually matter too much. In nLOS and NLOS links it matters a lot!
My thoughts, get you lower elevation side of the link up another 20 or 30ft. A good pop-up mast or a tall tree (threaded rod through the tree with large washers to a pipe is a good option as long as your tree is more than 5" thick at the rod placement) is a way to do this and the link planner install information includes final tilt angles that will help you get a good link.
Did some field testing of the link along the bridge path. The radios linked up at several spots along the path (where our road intersects to make these tests easier) closer to the mounted antenna - meaning within 1,000ft. But then I question the ability of me acquiring the link in some parts of the middle due to the height of the fresnel above the ground. Although that’s only the first fresnel (question about that at the end).
Strange finding: the link was in the -60s 200ft off the path, and better than on it:
Could that be scatter from the trees or a sign of the antenna quality?
Speaking of questionable quality. When I got back to the mounted antenna and parked, on the other side of the house from the mounted antenna (which points the other direction) I had a decent link. Makes me question if this truly is a PTP antenna. Or again, could this be from scatter off the trees?
Question regarding fresnel zones: I can’t find info on 2nd or 3rd fresnel zones. All I find is stuff on the primary one: Fresnel Zone Calculator - everything RF
It’s a good description through it doesnt get to far down the rabbit hole.
Your new signal issue could be scatter or a good reflection of the actual signal, that is difficult to determine remotely.
Your proposed link path is heavily dependent on accuracy of the endpoints gps location. Dont be surprised that your not on the path at one end. If you used a google map to get the gps coordinates then you have a certain level of error that is involved. Using a phone app requires you to switch to gps only mode to get accurate coordinates as cell tower telemetry requires three towers to be anywhere near close and that the towers gps coordinates be actually exact (not usually from what I have seen). Gps systems require a bit of time to make the initial calculations of enough satellites to work properly, the longer the better in my opinion. Turn it on, wait at least 5 mins then start walking around with it, make several passes past where you plan to mount and then stop at the mounting point, wait a minute so the things calculations catches up to your position and record the numbers. Link planner has the ability to take coordinates in just about any format so switch to the format you have the numbers in before you enter them, then select the format you want to use. Once you have a good proposed path calculated, check you AP end to make sure its within a couple degrees of the intended path. This is complicated as the closer to the metal of the antenna and mount, the more the inherent deflection of the compass will be.
Once you know your close to aligned horizontally then you can down tilt the AP antenna to match whats in the install report. Do the same for the SM but use the uptilt for that site.
A Yagi-Uda antenna has a bunch of elements called directors, and then has a ‘reflector’ at the back of it. BUT… Yagi’s are often transmitting half as much out the back as they are out the ‘front’… meaning that ‘reflector’ elements aren’t perfect. Add to that that Ubiquiti is famous for making “directional antennas which are so shitty that they are nearly omni” for many years.
But - to be fair - most antennas have some amount of side lobes and back lobes, so once you get close enough, it’s not uncommon to link through the back or side of the antennas. Better quality antennas will have a higher F/B ratio (Front to Back) and may have 30+ db more energy out the front, vs the back in a good antenna. And horns are even cleaner yet. But, 900 Mhz often uses Yagi, because the size of a parabolic dish or horn would be gigantic in such a low frequency.
Above is what they publish as the ‘pattern’ of the antenna. Most people on here would probably agree that Ubiquiti’s published specs are ‘optimistically inaccurate’ to say the least. BUT, even this optimistic drawing shows that there are lobes out the side and back which might only be 20 db lower power than the main beam. So… it’s very common to drive up 100 meters behind an antenna, and still get a link. Especially with a Yagi like this.
LINKPlanner does not provide a visualization of the second and third Fresnel zones, as this would make the path profile image too cluttered, but the impact is taken into account in the diffraction calculations and hence shown in the Excess Path Loss value.
It may be time to think outside the box and go a little off the deep-end.
Is that a lake in the Google Earth view?
If it is a lake and you are allowed to anchor long term why not setup a relay point by utilizing a small rowboat outfitted with radio, battery and solar panel.
Just think you might be nominated for most outrageous link of the year.
Your link is vertically challenged and platform shoes are not going to fix this problem. If this is something you really need and you will be utilizing it for a number of years buy a new or used tower. Go higher and look at other frequencies. Cambium has solid gear for almost any budget.
One other wack-a-doodle thing to consider is a Gigabit Media Converter.
Media converters and single mode fiber are inexpensive and can easily reach 2000 feet.
Get permission from the land owner to run 3/4 irrigation pipe or pvc conduit below the surface, thru the trees or on the ground. Better yet for less money get a few rolls of black poly tubing used for water supply lines. Black poly or irrigation tubing are low budget materials.
Using epoxy, attach a few metal nuts to one end of the fiber, then take a strong (rare earth) magnet and then pull the fiber thru the tubing.
If I recall there is a road intersecting your path. Get permission from whoever maintains/controls the road to cross it. That might be an expensive hurdle. You might luck out and discover a storm drain connecting both sides of the road.
I’ve actually done quite a bit of work with a geospatial analyst. We never used survey grade GPS (sub-cm) but we did use mapping grade GPS (sub-meter). To actually get sub-meter you have to stay at a point for at least 20 “readings” but really more like 50 for anything close to the actual position in the GPS estimate. And then you have to post process the data to really get to the equipment’s potential.
Its a pain, but nothing like survey grade or perhaps worse a differential GPS measurement (really good at elevation).
Still I think the best way is to plot the bridge path on a satellite image and then visualize where it is in the reality, as you said.
Funny stuff. I think I’d rather put a relay at a house on the other side of the estuary than a boat or even a pier in it. I doubt they have broadband over there.
Its pretty disappointing that I may be limited by sub-par (yet big name) Yagi antennas and/or the need to be a hundred feet above grade rather than the 20ft I thought was the deal. For comparison: I have a directional cellular antenna (passive) pointed towards a cell tower that improves my bandwidth 5x (or more - reaches 20mbps up/down). And that’s thru the trees and over hills to grandmother’s house, er, the nearest cell tower, at least 3 miles away. Hardest part was routing the cable into my house.
I know, its not the same thing (omni with tons of power) but I’m flagging. If I had to vent I’m annoyed that Cambium dropped their Yagi antennas, and a decent sized retailer of their equipment sold me some junk Yagi. Probably can’t sell these off either. sigh.
On top of that, I’m finding it difficult to motivate for these gesticulations when Starlink is approaching Alpha release. I’m not thrilled with their “better than nothing” beta release (they actually call it that), but for a $500 equipment charge and $100/month ($30 more than I’d pay for broadband over the wireless bridge) - with broadband latency/bandwidth - I’m tempted.
There you go. Maybe you could offer the person the ability to setup their own account at your location in exchange for space and power at their location for your relay.
If there are enough homes across the water without broadband you should draft a business plan to service them.
KP Performance are solid antennas. Poor choice in selecting the Ubiquiti antenna.
Don’t blame the equipment for what the terrain is causing. The math is what the math is. FWIW I’m grateful and blessed to have a 3.11 mile link working with one end 25 feet off the ground with 1.5 miles of trees in the middle.
It’s amazing what licensed spectrum and different tech can achieve. You are comparing apples to…
Didn’t know that Cambium stopped making their 900MHz yagi. But as others have said KP Performance makes a solid yagi.
If Cambium has stopped production of their yagi let’s hope they don’t do the same thing to their radio.
You needed two yagis and most likely they only had KP Performance in a 5 pack. So the salesperson did what a salesperson does and sold you the antennas they had. ISP Supplies sells individual KP Performance yagis.
If you are unable to build out the link on your budget ( money or sweat ) then having options are a good thing.
How much time did you spend on this forum asking questions prior to purchasing your gear?
The majority of people on this forum operate WISP. They have a lot of knowledge and openly share with the community.
I think you pulled the trigger too soon.
Today is a new day you have more information now then when you started.
