Digital EME on 2 meters

NOTE: This article is based on information and suggestions from Earl Shaffer, WB9UWA, who has been operating on 2m EME since 1999. I (KA1GT) have never operated on 2m EME. My experience is limited to 432MHz, 1296MHz and 10Ghz, where things may be quite different. Since I edited the article, all errors are mine! Don't blame Earl if I miss-interpreted information he provided, or made my own errors!

Doppler Issues

The maximum Doppler shift you will see on a 144Mhz signal reflected from the moon is about 350Hz. So if you (or a DX station) transmits on a frequency of 144.100000, the frequency of the signal when it comes back from the moon will be in the range 144.100350 to 144.000650. If you (or a DX station) tune to 144.100000 for TX and use a 1550Hz sync tone audio frequency, and you stay on 144.1000000 when receiving, the EME echo will show up somewhere between 1850Hz and 1150Hz on the widegraph (waterfall) display of WSJTX. This means you can get away with not using any Doppler correction or tracking since you will be able to see a reply to your call on the waterfall display. Many stations use no Doppler tracking, though of course if you want to you can use the CAT controlled Doppler tracking which is built into WSJTX.

Most used Digital Modes

Currently (1/2023), JT65B is far the most popular digital mode for 2 meter EME operation. Some stations (maybe 10%?) may use Q65-60A, but during contests, even they might default to using JT-65. Q65 has a number of advantages over JT65, and is used almost exclusively at frequencies of 1296MHz and up. However the difference in ultimate sensitivity between JT65 using Deep Search and Q65 (using AP search) is small (but is generally in the favor of Q65, at least on the higher bands). There is some disagreement about which mode works best under real conditions on 2m

This is not a complete tutorial on using WSJTX but rather a few key points that newcomers to 2 meter EME often miss. It is highly recommended that you read the key parts of the WSJT-X user guide which relate to general operation and setup of WSJTX and the JT65 section in particular.

JT65 can use an external database of stations known to be active on VHF/UHF frequencies and includes information such as their location and whether they are active on EME. This information is stored in a text file called Call3.txt. It's up to you to create this file, or to copy it from one of a number of websites which make it available. It is not automatically installed when you install WSJTX for the first time. Every instance of Call3.txt may be slightly different. There is no "standard" Call3.txt. When WSJTX is first downloaded and installed, Call3.txt is empty. It is important to find a current Call3.txt and copy it to the location in place of the empty Call3.txt. A fairly well populated Call3.txt can be download from a few web sites including LiveCQ and Make more miles on VHF. Call3.txt should be located in your WSJT log directory (File-> Open Log Directory). Deep search (Decode -> enable deep search) should be turned on to use Call3.txt which enhances sensitivity when using JT65. To test to see if a specific station is in Call3.txt enter their call in the DX call box. Click lookup and you should see his grid square. If not, likely he is new so you can enter his grid square and click add. The next time you enter his call in the DX call box and Click lookup, his grid square should appear and there will be enhanced sensitivity for his call provided deep search is on.

Most 2m EME stations use the N0UK EME logger to monitor band activity, You can also use LiveCQ to see which stations others are actually receiving. Some stations also use the HB9Q 2m logger (https://logger.hb9q.ch/).

WSJTX settings for JT65 operation

First you need good time synchronization, as you do far any digital mode. Your PC clock needs to be within 1 second of the correct time and it doesn't hurt (and may help) to be even closer to the correct time than that. The internal PC clock isn't very accurate so windows automates periodic re synchronization with one of a number of public time servers. Even that often isn't good enough and can result in significant time "steps" when synchronization occurs. Ideally you should install something like Meinberg NTP Client. There's more info in the WSJTX User Guide, but here's a short section from it:

"...The built-in Windows facility for time synchronization is usually not adequate. We recommend the program Meinberg NTP Client: see Network Time Protocol Setup for downloading and installation instructions. Recent versions of Windows 10 are now shipped with a more capable Internet time synchronization service that is suitable if configured appropriately. We do not recommend SNTP time setting tools or others that make periodic correction steps, WSJT-X requires that the PC clock be monotonically increasing and smoothly continuous...."

WSJTX Settings

Using the File->Setup->General menu, check decode after EME delay or you will never decode any EME signals! Set the Mode to JT65 and the sub-mode to "B" for 2m. Adjust sliders on the waterfall to give a good amount of speckles on the waterfall. I set the averaging to 15 to get high visual sensitivity. Make sure to check the sh box (sh = Shorthand). JT65b is almost always used with shorthand signals on 2m. Ftol (Frequency Tolerance) is normally 500 hz at its initial setting. This sets the range over which WSJTX looks for signals to decode. When you are calling CQ, you should set the TX sync tone at 1500 hz (though 1270Hz is sometimes uses). The TX frequency range is shown on the Wide Graph (waterfall) as a red slider. The green slider is centered on your receive frequency and can start at 1500 hz also but you want to move it when you notice a sync signal outside of its range. You do this by clicking on the waterfall at the sync frequency you want to receiving. You can get Ftol to 100 hz when you see a likely sync signal, though this isn't necessary if there are no other stations within its range.

You can clear averaging any time and it may take some experience and luck to get it right. It's not a bad plan to clear averaging at the start of your CQ and after every QSO. If you see a sync signal on the waterfall and there is no decode make sure the green slider is lined up over the sync line and hit decode. When you have your first decode, click on his call. That lines you up perfectly with the sync tone. At this point you have an answer to your CQ. He will be sending calls and his grid square. You respond by sending calls and "OOO". If he copied your "OOO" report, expect to see RO on your waterfall and at the decode time, an RO. Since you are lined up with his sync, you will see his shorthand signal on the waterfall both at his sync frequency and at the first green hash mark, indicating a visual decode of "RO". This is sufficient even if the decoder does not catch it. You respond by sending "RRR". He will see your sync tone and another line at the second hash mark. He should then be sending 73 and you will see lines at the first and last hash mark visually indicating a 73. Usually you will see the shorthand signals at greater sensitivity than the decoder, but rarely you will not see the shorthand signals, but they are decoded ok. The system is a few db more sensitive to shorthand signals so that allows for QSB during the QSO.

If signals were strong, clear averaging and call CQ again. If signals are weak or the QSO difficult then respond with your final 73. You should see that he stopped transmitting and you can call CQ again. Reset Ftol to 500 and see if you get another answer to your CQ. If a caller is strong, you wont have to set Ftol to 100 and you may find that you decoded several callers inside that window. Pick one and answer him. You can narrow Ftol to 100 hz on each trace and see who called you if they did not decode. Line up on their trace and click decode. You can do this for each station while you are answering a call. Give people a chance to find you. I try to call CQ for at least 20 minutes. Before calling CQ I announce it on the EME chat. Usually it is something like: CQ CQ CQ 144.144 2ND JT65b. Generally US stations transmit second. This helps stations in Europe handle QRM. When you are a new station, I recommend saving time by calling CQ. Make sure others know you are calling and as they copy you, they will call. QSB is such that different parts of the world will peak up at different times so you could be wasting time doing a sked with only one station that is in the wrong part of the world at that moment. For your first try, see if you can copy a few stations, then make a sked with one of the larger stations or answer their CQ. Once you have established that you can work EME, its then time to call CQ. If you did a good job with the hardware and picked a good time to operate, you may find yourself at the business end of a pile-up. Everyone is anxious to work an initial.

2m EME Equipment

A full description of the hardware required for EME is beyond the scope of this article so I'll try to touch on things as they apply to 2 meter EME that may not be obvious to the new EMEer. First of all, try what you already have. If you have a single yagi set up for tropo and at least several hundred Watts, get on the air and try horizon moon shots. A lot of EMEers start this way and is one of the draws to 2 meters. Many stations never put up a full AZ/EL EME system but work EME on the horizon. I have worked many of these stations. There is nearly always someone new to work. If you don't already have a 2 meter array, then try a single yagi. Put it up about 6 meters above the ground pointed at the horizon and try for ground gain toward the moon. About 15 degrees moon elevation may be optimum. Usually several hundred Watts is needed but 100 watt stations have worked some EME station. As you might suspect, more power yields more contacts. You should shoot for 1.5KW. Some stations run much more than that and have proportionally better results. It is generally easier to be copied off the moon than to receive EME signals. This is mostly due to noise. If you can do EME from a rural location, then that is preferable. If you are stuck in an Urban location as I am, you really need the lowest sidelobe array that you can find or build. That being said, I recommend 4 yagi arrays for everyone. In an urban location a two yagi array may work but likely you will need a minimum of 45 degrees moon elevation to have good RX. I recommend an LFA yagi. My home brew yagis are designed and built by me and they have lower sidelobes than the LFA and my sidelobes are often too high for my large noise environment. Until antenna manufacturers put more emphasis on lower sidelobes the LFA is likely about the best one can do with commercially available antennas.

Longer yagis will help. Four yagis are far better than two and Xpol is better than single polarity arrays unless the array is very large, such as 8, 5wl yagis. Even in rural locations noise is an increasingly big problem so the LFA design is still very desirable in that environment. Likely a 2 yagi array will work ok in a rural location, but you probably have room for a 4 yagi array in that location so why not? You can make a lot of contacts with a single polarity so you don't need Xpol, at least initially. Xpol generally results in faster QSO's so it is a great choice for working contests. I won't go into detail here on Xpol because there are so many ways to set it up but I would like to say that you should plan on transmitting one polarity or the other and NOT transmit circular polarity unless you have a large station with decent power. If you have less than 1.5KW don't even think about transmitting circular. In theory, you can transmit circular and you will not have to ever switch TX polarity, but you are limiting your TX ERP and there are some stations that will never copy you or it will take far longer for them to copy you. Faraday alters polarity and it gets a lot of credit for QSB, but there are other sources of QSB and that can easily be 6db. If you think I am trying to discourage the use of circular polarization on 2 meters, you are correct. Your TX signal will be 3db down from what it could be with matched polarity. You will often try to work stations that are pointed into their noisy horizon so when you throw away 3db of TX power, you may never reach a point that you overcome his noise level.

Your station should have an array mounted LNA. The LNA should have a noise figure below .5db and moderate selectivity and IP3. It needs to reject the FM BCB. An LNA with a wide open selectivity should not be used here. If you already have an array that has several bands in close proximity, by all means try it on EME. Just know that your nearby yagis for other bands will degrade the pattern and to a lesser degree, the gain of your array. The result will be that you may be copied off the moon and the pattern degradation will make receive more difficult. The end result is that your multiband array will perform on par with a much smaller single band array. There are some compromises that are possible when trying two bands on the same mount, but I recommend getting 2 meters going by itself first and once you have a good idea of how it is working, add the second band and see if 2 meters is still ok. Better yet is to model it first. Spend some time measuring noise. It should be done with no AGC. It's a good idea to set up the LNA so its input can be terminated by a 50 ohm load. You can then directly tell how much noise is added by the array. Linrad with an SDR input is a great way to keep track of noise. If your array only adds 3db of noise, you are in good shape. 10db more noise will make EME very difficult.

The array should be a low as practical to minimize noise. You can use a short yagi to find noise sources and possibly locate a good spot for the array. If a single tree blocks part of your moon window, you are probably still in good shape. End mounted arrays nearly on the ground can be very effective in a noisy location. In a rural location one could set up a dream horizon station. With flat open terrain in front of the array, 4 yagis could be set up side by side on a 30 foot tower with a lift on it. The height of the array could then be set up to extend operating time by changing height to move ground gain peaks and nulls. Ground gain can be as high as 5db, but ground reflection can work against you too. 4 yagis side by side on a lift would be easy for an old man to maintain and of course it could be set up for elevation too. Parking the array at a low height would give it good wind survivability. An obstructed view of the horizon is likely to reduce ground gain to zero. That's what I have with houses all around me in close quarters. I count myself lucky if I only have a 20 db noise rise in any direction on the horizon. Good DF equipment for noise and good PR is a key to success in an urban environment.

Content by Earl Shaffer, WB9UWA, 2 METER EME since 1999.