Starting EME: what 2 m, 70 cm, 23 cm and 10 GHz moonbounce really demand (JT65/Q65, Faraday, libration, skeds)
- Replies 2
- Views 35
- Created
- Last Reply
Top Posters In This Topic
-
Admin_Lee 1 post
-
Mod_Chrissy 1 post
Moonbounce has a reputation for needing a huge station, but the requirement changes a lot by band. These notes are a starting point for comparing the bands and for deciding where to put effort first. Numbers are ballpark figures. Check them with a link-budget tool before you build anything.
Path loss is the starting point. The round-trip loss to the Moon and back is roughly 252 dB on 144 MHz, about 261 dB on 432 MHz, about 271 dB on 1296 MHz and close to 289 dB at 10 GHz. The Moon reflects only a small fraction of the energy that hits it, and the distance swings by about 2 dB or a little more between apogee and perigee. Each band up costs you about 9-10 dB in path loss. You get most of that back from antenna gain, because a fixed physical aperture gains more as the wavelength shrinks. Sky noise matters too, which is why the best dates and times for a given band are not always the same.
On 2 m the usual entry point is a stacked array of long Yagis, commonly four, at around 17-19 dBd, with a high-power amplifier. Beamwidth is wide, in the region of 10-20 degrees, so tracking is forgiving. The Moon has to be well above the horizon, because low elevation brings ground noise and spill-over. Galactic background noise is high on 2 m and varies a great deal depending on where the Moon is in the sky. Faraday rotation is the other big factor on this band (see below). On 70 cm, arrays of several or many Yagis are common, and so are large dishes. On 23 cm the usual choice is a dish, with 3 m or larger being a typical starting size. A 3 m dish gives roughly 29-30 dBi at 1296 MHz with a beamwidth of about 5 degrees. At 10 GHz a 1.8 m dish gives roughly 43 dBi and a beamwidth under 1 degree, so pointing accuracy becomes the main problem. Even a modest dish at 10 GHz needs a tracking system that stays within a fraction of a degree.
Receive sensitivity is worth as much as transmit power. Put a low-noise preamp at the antenna, with a noise figure well under 0.5 dB on the lower bands, and keep the loss ahead of it as small as you can. Use a good low-loss relay and a proper sequencer so the preamp is protected and nothing hot-switches. Transmit-side feedline loss costs you directly, which is why people fit rigid or low-loss line and put amplifiers close to the antenna on the higher bands. For dishes, Sun noise and Moon noise measurements are the standard way to see whether the feed, preamp and dish are working as they should. If they disagree badly with the numbers you expect from the dish size, find out why before blaming the sky.
Polarization is where the bands differ most. Faraday rotation twists linear polarization as the signal passes through the ionosphere, and the effect scales roughly as 1/f squared. On 144 MHz it can cause deep, slow fading. It is much weaker at 432 MHz and negligible from 1296 MHz upward. Spatial polarization offset is a separate geometric effect. Two stations at different places can see the Moon's polarization plane rotated relative to each other, even without Faraday. This is why 2 m stations often use crossed Yagis with switchable or adjustable polarization, and why a partner who looks weak may simply be misaligned. On 23 cm and above it is common to use circular polarization. The reflection reverses the sense, so the usual convention is to transmit RHCP and receive LHCP, which works at both ends.
Libration fading is a different effect. The Moon's apparent rocking produces a spread of Doppler shifts, which shows up as rapid flutter on the signal. It scales with frequency, running from a fraction of a Hz on 2 m to a few Hz on 23 cm and more at 10 GHz. This is why digital modes for the higher bands use wider tone spacing. In JT65 you pick submode A on 144 MHz and the wider B or C submodes for 432 MHz and up. Q65 has A through E submodes for the same reason, and the 60-second period versions are the usual EME choice. Q65 was introduced in WSJT-X 2.4.0 and handles Doppler spread better than JT65 does.
Doppler shift from Earth's rotation is also large, about a few hundred Hz on 2 m, several kHz on 23 cm and tens of kHz at 10 GHz, so most operators use automatic Doppler correction. WSJT-X has a built-in Echo mode and an Astronomical Data window. Together they let you check pointing, frequency and your own echoes before you try to work anyone. Your PC clock also needs to be accurate to well under a second, through GPS or NTP.
Skeds follow the same logic as the rest of the path. You need a window when the Moon is above the horizon at both ends, ideally at a good elevation, with acceptable sky noise. Tools such as MoonSked, VK3UM's EME calculator and the HB9Q logger page help with scheduling and with estimating whether a path is workable. The usual convention is that one station takes the first minute period and the other takes the second. Confirm the arrangement in the sked, because the assumptions differ between groups.
Questions for the thread: what is the smallest station you think can realistically make JT65 or Q65 contacts on each band? Which of the above has been the biggest obstacle for you: tracking, noise figure, polarization or schedules?
Link to comment
https://www.hamradiobase.com/forums/topic/10941-starting-eme-what-2-m-70-cm-23-cm-and-10-ghz-moonbounce-really-demand-jt65q65-faraday-libration-skeds/Share on other sites