Building toward EME: how band, array size, mode and polarization decisions fit together (2 m to 10 GHz)
- Replies 2
- Views 56
- Created
- Last Reply
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
EME planning goes wrong when people size the antenna first and ask about the rest later. This thread is meant to line up the main variables so we can compare approaches. Corrections and different views are welcome, especially on array sizes and band choice, which vary a lot by site.
Path loss and what it means per band. Round-trip EME loss is roughly 250 dB on 144 MHz, about 260 dB on 432 MHz, and about 270 dB on 1296 MHz. It keeps rising with frequency, though antenna gain also rises for a given physical size. Distance to the Moon changes between perigee and apogee, which costs about 2 dB of round-trip signal. Perigee is an easy way to pick better days for a marginal station. The echo delay is around 2.5 seconds, so you can hear your own echo only after you stop transmitting. That delay is also how you check your own system.
Band by band, in rough terms. On 2 m, sky noise is high and antennas are physically large, but the Doppler spread is small and the power amplifiers are comparatively easy. Typical stations use stacked long Yagis, often four or more, with serious legal-limit power for the weaker setups. On 70 cm, arrays tend to be bigger in element count but smaller in boom length. Faraday rotation is much less of a problem. On 23 cm, a dish of about 3 m with a good feed gives roughly 30 dBi. It also gives a beamwidth of several degrees, so tracking is manageable. On 10 GHz, a dish of the same size gives gain in the high 40s dBi, but the beam is well under a degree. Pointing, mechanical stiffness and a good moon-tracking routine matter much more. Treat all of these numbers as rough. Feed efficiency, surface accuracy and feedline loss will move them.
Receive side. Every 0.1 dB of loss before the preamp adds roughly 7 K of noise. This is why the low-noise amplifier should sit at the feed or antenna, and why the TX/RX relay and any connectors ahead of it deserve attention. At 144 MHz, sky noise is large, and it is even larger toward the galactic plane, so squeezing the last 0.1 dB from the preamp matters less than choosing times when the Moon is in a quieter part of the sky. At 1296 MHz and above, a very low noise figure and a dish with low spillover pay off more. A sequencer for the preamp, relay and PA is not optional at EME power levels. Switching at the wrong moment can destroy a preamp.
Modes. JT65 was long the standard digital EME mode. JT65B is used on 144 and 432 MHz, and JT65C on 1296 MHz and up. The sub-modes differ in tone spacing, which roughly doubles per step (A to B to C). Q65, in current WSJT-X versions, has similar sub-mode steps. It also includes averaging over multiple periods and better handling of spread signals. Lower bands usually use the narrower sub-modes, and microwave stations go wider because Doppler spread grows with frequency. Check the current WSJT-X guide and the activity groups for the sub-mode in use on each band, because conventions can change.
Doppler and libration. Doppler shift from Earth's rotation can reach a few hundred Hz on 144 MHz. It is about ten times larger at 1296 MHz and into the tens of kHz at 10 GHz. WSJT-X can apply Doppler correction, so both stations can aim for a fixed frequency at the Moon. Libration is different. It is the Moon's apparent rocking, which spreads the echo over a range of frequencies. The effect gives fast, flutter-like fading and gets worse as frequency rises. On microwaves, the echo is spread widely enough that a narrow mode with tight tone spacing can fail even when the signal is strong.
Polarization. On 144 MHz, Faraday rotation in the ionosphere can rotate a linear signal by large angles. This gives slow, deep fades over minutes. Rotation scales roughly with 1/f², so on 432 MHz it is far smaller, and by 1296 MHz and above it is generally negligible. Separately, there is a geometric spatial offset between two stations that depends on where they are relative to the Moon. Many 2 m stations deal with these with switchable or rotatable polarization, or with dual-polarization arrays. At 1296 MHz and above, circular polarization is the norm. A reflection reverses the sense of circular polarization, so your own echo returns in the opposite sense. Agree the convention before a sked and do not assume it.
Sked planning. A sked needs the following settled: band, mode and sub-mode, nominal frequency and Doppler handling, who transmits in the first period and who in the second, start time and duration, and a fallback plan. Check that the Moon is above a usable elevation at both ends for the whole sked. Also check declination, sky temperature behind the Moon, and perigee or apogee. The WSJT-X Astronomical Data window shows many of these values, such as azimuth, elevation, Doppler, sky degradation and spatial polarization offset. Exactly which fields appear depends on the version. A rising or setting Moon at one site brings extra ground noise and refraction, so consider that when picking times.
Questions for the thread. What array size and power have people found a practical entry point on 2 m with Q65? How do you decide between 432 and 1296 for a first dish or array project? And what is your approach to polarization control: switched arrays, rotation or just waiting out the fades?
Link to comment
https://www.hamradiobase.com/forums/topic/10775-building-toward-eme-how-band-array-size-mode-and-polarization-decisions-fit-together-2-m-to-10-ghz/Share on other sites