That does represent a limiting case but it's a bit pessimistic. The longest path is for very shallow incident angles eg a 3 mm thick and 150 mm radius disk gives an angle of only about 1 degree. At 10 degrees, the path is about 20 mm; with a refractive index of 1.5, the path is only 10 mm longer. 10 degrees might be what you set in the receiver's elevation mask.
A mm thick layer of dirt is just going to be roughly another mm of plastic; worse if it's absorbed moisture, true. You have a point about water. Water has about 10 times the refractive index of plastic (real part of n) so this is more of a worry. A 1 mm film will have triple the path through a nominal 3 mm of plastic so at 10 degrees incidence there is now about 40 mm extra path which you might see in post-processing. But you're not going to see that in the 1 pps. Cheers Michael On Wed, 7 Feb 2018 at 8:14 am, Bob kb8tq <kb...@n1k.org> wrote: > Hi > > Since we are talking about an L1 / L2 antenna here, a reasonable assumption > would be that the target is something better than an “average result”. If > you construct > a cover out of a piece of PVC pipe (as shown in the original drawing), > your worst > case path has a foot or so of PVC in it compared to a best case path with > well under > a tenth of an inch. That’s going to give you a bit of variation ….. Add > some dirt or water > or ice to the equation and who knows what the result might be. > > Bob > > > On Feb 6, 2018, at 3:45 PM, Michael Wouters <michaeljwout...@gmail.com> > wrote: > > > > I can see why the geodetic community would worry about antenna phase > centre > > variation when a radome is installed but is it really an issue in timing > > applications? The few papers I've read suggest PCVs of less than 10 mm, > or > > equivalently, 30 ps. This is at the level of precision available from > > post-processed, carrier phase time-transfer but invisible in the 1 pps > > coming out of your receiver, even with a good sawtooth correction. Am I > > missing something? > > > > Cheers > > Michael > > > > On Wed, 7 Feb 2018 at 4:14 am, Bob kb8tq <kb...@n1k.org> wrote: > > > >> Hi > >> > >> There are “cell site” specific GPS antennas on the market. Panasonic has > >> had one out > >> for quite a while. I’m sure there are several others. > >> > >> One issue with doing any sort of “cover” for a precision antenna is > >> distorting it’s pattern. > >> Plastic (or whatever you use) will have different properties than air. A > >> path through a blob > >> of “not air” will change the effective path length. That impacts the > >> timing and thus the > >> navigation solution. If you are worried about 2mm sort of pattern > >> accuracy, things get > >> tricky. Early on, there was a big “throw out the radomes push when this > >> was first noticed. > >> > >> Bob > >> > >>> On Feb 6, 2018, at 6:15 AM, Bo Hansen <timen...@rudius.net> wrote: > >>> > >>> Hi > >>> > >>> Besides the RF characteristics it may also be worth considering the > >> quality of the plastics used. Over time water ingress may become an > issue. > >> Fours years after the installation of a CN brand antenna, sourced > locally > >> so probably not counterfeit either, we had to replace it at OZ7IGY > >> www.oz7igy.dk > >>> > >>> RF wise 42 dB of gain IS an issue. Again at OZ7IGY, with 12 carriers in > >> the air especially 13 cm and 23 cm, blocking and IMD were an issue > before > >> we mounted a BPF. I have taken apart the above mentioned antenna, a > >> Motorola antenna and an eBay "hockey puck" antenna. The best design was > >> clearly the Motorola one because it had a BPF after the pre-amp - > probably > >> because it was designed by RF competent people too. Each of the other > ones > >> had two FETs/MMICs in series and then a BPF. Of cause if no nearby > carriers > >> are in the air it may be less of an issue. > >>> > >>> So designing a really good antenna and pre-amp may be a business > >> opportunity. There are many hi IP3 MMICs available designed for GPS and > the > >> like purposes. SAW BPFs with <1 dB loss are available fairly cheap so > one > >> before the FET/MMIC with a 1 dB NF is the way to go. A DIY radome using > >> standard materials from any hardware shop is attached. > >>> > >>> Bo, OZ2M > >>> > >>> <GPSAnt_DIY_Radome.png>_______________________________________________ > >>> time-nuts mailing list -- time-nuts@febo.com > >>> To unsubscribe, go to > >> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > >>> and follow the instructions there. > >> > >> _______________________________________________ > >> time-nuts mailing list -- time-nuts@febo.com > >> To unsubscribe, go to > >> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > >> and follow the instructions there. > >> > > _______________________________________________ > > time-nuts mailing list -- time-nuts@febo.com > > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > > and follow the instructions there. > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.