Electronics > Projects, Designs, and Technical Stuff

Advice for roof-mounting GNSS antenna

(1/3) > >>

I'll be using a Novatel GPS-703-GGG pinwheel GNSS antenna (https://novatel.com/support/previous-generation-products-drop-down/previous-generation-products/gps-703-ggg-antenna) for precision timing applications. I'm planning to mount it to my roof and, as I've never done this before, would appreciate suggestions for achieving best results.

I've attached pictures of the roof I'll be working with. I suppose the first consideration is roof location. The easiest location for me to access is the lowest section of the roof (in image 3). This would also leave the cable to the GNSS receiver the shortest. But, maybe I should be concerned with reflections off the higher roof segments at this location. Another possible location would be the lower left corner in image 2. That's a bit further from the window (i.e., longer cable) but should have a slightly clearer sky view. Finally, I could try to mount it near the top for the best sky view, but that would require clambering up the roof, which I'm not eager to do. It's worth noting that the pinwheel antenna should provide pretty good multipath rejection.

Now for mounting. The cinder block method used by sparkfun (https://learn.sparkfun.com/tutorials/how-to-build-a-diy-gnss-reference-station/all#introduction under "Affix Your Antenna") seems like a good, cheap option. My roof does have a bit of a grade to it - about 13deg at the lower sections (I didn't measure the steeper grade of the upper section). If I need to be worried about the cinder block sliding off, I could drive in some nails downhill of it (other suggestions welcome too!). Another, more expensive option, would be to use something like the Pasternack non-penetrating antenna mount (https://www.pasternack.com/non-penetrating-antenna-peak-roof-mount-1-pole-version-60-inch-mast-galvanized-steel-with-powder-coating-peprm1-p.aspx). This has a 5ft mast. This would use cinder blocks to weigh down the mount, or drill holes with screws for fastening. How much benefit, if any, would there be in having the antenna mounted further off the roof? Another option, which would require drilling screws into the roof, would be something like this: https://www.ebay.com/itm/124845460270. That still mounts the antenna closer to the roof but is cheap and I wouldn't need to worry about a cinder block sliding off the roof and crushing someone below or breaking my antenna and pulling my GNSS receiver out of my window... Any special considerations for drilling into a roof?

I'd probably go for the last mounting option (short, screw mount) with the easiest location (3rd picture), unless people recommend I do something else. Thoughts?

In case it's relevant/useful, I'm in Oakland CA. I'd been planning to mount the antenna in the direction normal to the roof, since that's easiest to mount. With the slant, that would have the antenna tilted a bit north (actually about 15deg east of north). But, if it's better to have the antenna facing straight up or potentially even tilted in a different direction, I'd be happy to consider that too.

The novatel antenna has a 5/8"x11 thread.

The "Birds" are far out in space, so a few meters of extra height doesn't matter.
What matters is view ... Best is 360 degrees , and no roof blocking.

Reasonable cable length (10..30m) usually don't matter if you use a decent cable.

I'm using a good quad shielded 75ohm cable (Sattelite cable)  on mine , and even use F-connectors. As they are so easy to fit.
The impedance diff between 50ohm & 75 ohm receiving, on that length of cable doesn't seem to matter.

I have an N-F adapter on my timing "Ice-Cone" , and another N-F on my GPS Splitter.

Re mounting: I'd just use some of the Antenna mounting brackets , or even a Sattelite dish mount.
Just had a look on the antenna now, seems like it's using the "survey pin mount"
Something like this would prob. fit

Maybe a aluminium would be better (doesn't sway)


If you have lightning in your area, be SURE to provide a stout ground to the coax.


--- Quote from: bingo600 on June 09, 2023, 12:45:07 pm ---What matters is view ... Best is 360 degrees , and no roof blocking.

--- End quote ---

There's a flat section on the top of the roof. That should provide a pretty good, 360deg horizon to horizon view, and should be large enough to install multiple antennas with adequate spacing. My house is one of the tallest in the area, which should help. It's around the same height as the tallest surrounding trees too. I'll have to get up there though. The roof is asphalt shingle and provides a good grip. Plus, the grade isn't too steep. I can install a roof harness as extra protection and I'll probably just leave it up there so maintenance is easy.

--- Quote from: jmelson on June 09, 2023, 02:43:04 pm ---If you have lightning in your area, be SURE to provide a stout ground to the coax.

--- End quote ---

Yeah, lightning protection is definitely something I plan to incorporate. I'll use coax with good shielding/ground anyway. From a quick look, it seems that the common solution is to use an inline lightning arrestor outside the house with a good ground connection to earth ground. Is it enough for me to connect this ground with suitable AWG wire (eg 6 AWG or less) to an ground prong outlet on the outside of the house, or should I try to track down the earth ground rod and attach it to that somehow? Is this enough protection? Anything else I should do? Lightning is pretty uncommon where I live, but I really don't want an incident frying all my lab equipment.

The best protection will be coax to ground level outside of the structure, to the common structure ground, where the protection device should also be located. If this is at a separate ground rod from the structure ground system,  then the two should be bonded together. From that protection device, ingress to the structure and to your equipment location. That is the right way to do it. Bonus points if you then have a common signal and AC ground point at the room before it goes off to individual gear AC or signal connections.

Or you can just do what many people do and throw out those best practices, running a ground wire from your antenna to ground and believe you are protected, but actually be worse off than if you didn't have any ground. I would avoid this. There will be pushback to my statement here, as yes many satellite installers do it this way, and it is maybe acceptable to some, but I don't like it because you can end up with different ground potentials in your system, and those surges easily take out equipment as the potentials equalize through the equipment to AC ground.

FYI, some insurance companies will not cover if you do not follow the best practices. YMMV


[0] Message Index

[#] Next page

There was an error while thanking
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod