Table of Contents
Front Page | Antennas | Hardware | Software | General Tips | Gallery
Antenna considerations + links to projects
V dipoles
(Will add links to some good quality construction plans/ideas here.)
Easy to make, scope for modest improvements, but annoyingly low gain at times. You may have to tilt your antenna from side to side (so not facing straight up) to aim the pattern at a satellite. Try to visualise the waves coming out from the satellite, but also that you’re not directly underneath it and on a curved surface. Then it’ll make sense. Maybe. Or make a QFH.
The altitude of deployment will make a difference to your receive plots, depending on whether you’re using the actual ground as a ground, or reflectors (see below). Essentially higher up gives good low-angle gain, low down and reflectors give good overhead/mid-angle gain and shape the pattern more that way to cover any nulls. It may be beneficial to raise the antenna at the start and end of passes if you are able to.
Depending on how well you’ve made your antenna you may have some directionality to it, and may receive a better signal at a totally unintuitive deployment angle. Of course, when the satellite is close to you the angle matters less and the received signal power increases.
Clip-on ferrites can be useful to reduce static/noise (check the specifications first!), and some people use air chokes with the coax around a pole. You could also use a proper balun, if you liked, but there’s probably no need to.
Notes on Reflectors
The simplest reflector is an identical (or slightly longer) set of elements about a 1/4 wavelength below the main connected element. This may kill your low-angle gain depending on your LOS and the particular satellite, but will help your overhead gain resulting from changes in polarisation throughout the pass as the satellite tumbles and moves in its orbit. When MM3IIG tried it there wasn’t much difference in low-angle gain as long as you have decent LOS, and you’ll still pick up signals at around 15-20 degrees of elevation with enough SNR to decode.
Also, the number of reflectors can make a difference. In a particularly detailed Reddit post a user went for six of them 60 degrees apart each, but for most people will get away with a single pair of elements underneath. Interestingly, if you go for a circle it seems you don’t angle two of them in line with the connected element above.
Other resources online
Here’s a fantastic video that goes through all of the different combinations of mounting height and reflectors, with actual plots and data. If you’re set on a V and scared of making a QFH this will help (Anirudhha Bokil VU3FML): https://youtu.be/0x2JTjgGyAY
And here’s some Reddit posts with great comments on construction, reflectors, and other topics:
- Good run-through on the various possible designs - like the video above but in article form: https://www.reddit.com/r/amateursatellites/comments/cnnjjm/how_i_came_up_with_my_current_vdipole_design_and/
- A guide on making a 6 reflector design: https://www.reddit.com/r/RTLSDR/comments/c1l6l0/my_best_vdipole_so_far/
Turnstiles
Why are these used if dipoles are pretty okay but QFH is the best?
QFHs
Discussion on polarisation and angles, why this is the best option
Find the easiest plans and link here
Club build gallery (Simon and John)
QFH Calculator: http://jcoppens.com/ant/qfh/calc.en.php
Other options
Double cross/double dipole
Any others?