NOTE: This article was written before FT-8, and even MSK144, became popular. Many of the principals I mention are applicable to both, where I mention FSK441 then MSK144 can be substituted, and where JT6m/JTMS etc is mentioned then FT-8 can be applied. I make regular updates (when time and enthusiasm allows) so contact me if anything doesn’t make sense or you think some information is inarticulate or unclear.
I’d like to share with you all my passion for working stations on VHF (6m, 4m and 2m) using Meteor Scatter (MS) as the propagation mode. I wrote this during the peak of the Perseid meteor shower in 2015 as it’s much more likely that a newcomer, even with a modest VHF station, inspired by the article to try and decode MS bursts, will decode bursts during the meteor shower peak, which in turn hopefully increasing the chances of them continuing to try MS after the peak has died down in the knowledge decoding is possible. Even though the shower I wrote this for is long gone, the article is still a good guide to get you started in MS.
But before we start, can I please encourage everyone not to hold MGM QSOs on the MGM calling frequency for the band/mode you are using, not including QSOs using FT-8 of course. Anyone who CQs on an MGM calling frequency using modes like MSK144 or FSK441 should do so with a QSY frequency (a form of split operating where you TX CQ on the MGM calling frequency but RX on an adjacent frequency which you then QSY to to complete a QSO if your CQ is answered). Operating this way allows others to use the same MGM calling frequency at the same time as you, if you CQ with a QSY, then your RX is not blocked by a local station calling CQ on the MGM calling frequency in the opposite period to you, as even in the UK a station could be CQ’ing in the 1st or 2nd period.
If you hear a local station who is not CQ’ing with a QSY, I encourage you ask them to call with a QSY. If you decode a [DX] station via an MS ping then see if they are on ON4KST chat and ask the station to complete continue the QSO on an adjacent frequency. I only ask this for the benefit of all stations who wish to have fun making QSOs via meteor scatter, I feel disheartened when I see stations operating simplex on any calling frequency, it is in my mind poor, even arrogant, operating.
By a modest VHF station I really do mean a simple station like an FT-817 or SDR as the receiver and a vertical collinear as the antenna. Of course a better radio and antennas like a beam would be better but it’s much easier to get started than many think, and that’s what I want to show. This has been proved during the writing of this article where a friend Stevie, 2M0ZFG in IO76, surrounded by 500m hills all around him (see his horizon plot below) and using an FT-817 radio + V2000 vertical collinear antenna (fed with RG-58) has been decoding MS pings on 2m (FSK441) from various EU stations and Steve, M0MVB, who has been decoding EA2ARD on 6m (JT6m) using his Kenwood TS-2000 radio and V2000 vertical co-linear antenna.
Image courtesy of HeyWhatsThat
Quick Meteor Scatter (MS) Overview
Simply put, MS QSOs are made by reflecting (bouncing) your signal off the ionised trail of a meteor as it burns up when entering the atmosphere.
Image courtesy of QST
Meteor trails are unpredictable and “random”, however there are times when the number of meteor trails increase, these are early morning, and during meteor showers, it’s during these times that your chances of making a QSO using MS as a propagation mode are at their highest. The increase in the meteor count during showers is obvious but the reason why early mornings also have an increase might not be so. It’s down to the fact that the point on the earth which has the sun rising, your location if it’s early morning, is facing forward in the direction of travel of the earth through space and so collides with more debris, but this also means that from noon till midnight the opposite is true, and it is the worst time of day for meteor reflections as your point on the planet has moved away from being at the front and is being somewhat shielded from space debris.
Image courtesy of G3WZT
To get started in MS you just need to do the following (for those starting with VHF meteor scatter, I suggest using MSHV, but other software like WSJT-X is an option);
1. Ensure your PC clock is accurate, see the “PC Timing” section below.
2. Download MSHV (link below).
3. As a minimum to RX, configure the sound card settings (Options > Sound Settings).
4. Tune your radio to the correct frequency and mode (see the band/mode table below).
5. Click on ‘START MONITOR’ in MSHV.
6. Sit back and watch (and most likely wait).
7. When you are ready to TX, hopefully after you have spent some time monitoring the activity of other stations and their QSOs (so you have an appreciation and understand of the MS/MGM procedures), see the “Points On Tx” below, and take note of my encouragement at the start of the article.
MSHV in more detail
Decoding MS pings needs a receiver capable of receiving [the mode] USB, a computer with a sound card and an interface of some kind to connect the audio from the radio to the sound card. There are 2x software packages available to decode the MGM modes mentioned in this article, which include FSK441, MSK144, JT6m and ISCAT (more on these below). These packages are WSJT-X and MSHV by LZ2HV, MSHV is the simplest software package to setup and use. With mentioning is JTDX (which is aimed at HF/6m users) and the older WSJT (non ‘-X’ versions).
Download and install MSHV – it’s easy to configure & LZ2HV has a guide on his website if you struggle. Remember, if all you plan on doing is listening for meteor bursts then all you need to configure are the sound card settings and RX mode, you don’t need to change the callsign, macros or PTT settings.
Configure the sound card under the ‘Options’ menu;
Then click on ‘START MONITOR’ to start MSHV listening for MS bursts;
Unpredictability of the Meteor (when to TX)
Due to the unpredictability of meteor pings, a simple convention is used which maximises the chance of a station hearing you if you are calling CQ randomly. The convention is this;
- Any station beaming West transmits in the 1st period.
- Any station beaming East transmits in the 2nd period.
In region 1 (Europe, Africa, the Middle East and Northern Asia), the convention is only a guide, not a rule, but do your best to follow the convention, even if others are not, when you call CQ.
This of course is easy to follow if you are using a directional antenna, but given the popularity of FT-8 on VHF, I know many stations start by just using omni-directional vertical antennas like a V2000. This leaves a question regarding which period to transmit in as you can’t beam East or West with an omni-directional vertical. To stay best aligned with the transmit period convention above, I suggest the following idea should be applied, especially if you are calling CQ;
- If you are beaming or looking to work stations who are West of you, then transmit in the 1st period.
- If you are beaming or looking to work stations who are East of you, then transmit in the 2nd period.
When operating on FT-8, keeping to the convention can be hard, especially during a Sporadic-E opening when using a non-directional antenna, say a V2000 vertical. However, please be mindful that other stations might be trying to hear or work weak DX (that you might not be hearing) and if you don’t stick to the convention then you potentially risk causing undue interference to those other stations. It’s also entirely possible a DX station is not following the correct convention and is transmitting in the wrong period , so keep an eye on your decodes to see if that is the case and if so act accordingly, which might mean you change transmit periods, or stop transmitting for a few periods to allow the other stations their chance to work the DX station, believe me, they’ll appreciate it.
Finally, in the unlikely event that the direction between stations is within a few degrees from due North or due South, then use best efforts regarding the period to transmit in, again being mindful of others.
1st or 2nd Period, 15 or 30 seconds?
The period length can change dependent on the band / mode used. On 6m + 4m 15 second periods are used, and on 2m 30 second periods are used. If the period is 30 seconds it means the 1st period is between 00-29 seconds (of every minute) and the 2nd period is 30-59 seconds, the result is the station who is more East (beaming or looking West) transmits between 00 and 29 seconds and listens from 30 to 59 seconds, while the station most West does the opposite. If using 15 second periods then the 1st period uses 00-14 & 30-44 seconds, and the 2nd period uses 15-29 & 45-59 seconds in each minute. These defined periods were created before the days of internet chat rooms like ON4KST when most MS QSOs were random.
When using an omni-directional use the period you like but if your intention is to work stations to the east of you remember that are likely TX’ing in the 1st period so are therefore listening in the 2nd. If you have a sked (a planned tune and place for a QSO) the relative position of the station to you (more East or West of you) should determine the period, you then set the period in the software and it takes care of the rest. When using the software in RX only (called Monitor) it will decode both periods allowing you to see all the activity.
The timing of the PC clock is important for all this to work, otherwise you and your QSO partner will be out of and you’ll both end up transmitting at the same time for however long your clock is out. Lets assume your PC clock is 10 seconds slow, there are 2 consequences to this, one practical and one social,
- It will mean for 10 seconds in every period you’ll be TX’ing when you should be RX’ing and RX’ing when you should be TX’ing, ultimately reducing your chance of completing the QSO by 33%.
- Also as your period is 10 seconds late in effect, you’ll end up TX’ing outside the expected period which will cause QRM to local stations who are potentially listening for DX
To solve this you need to regularly synchronise your PC clock. Regularly is not every 24 hours but every 15 minutes. The simplest way to do this is to install a piece of software called Dimension 4 which “is the fastest and easiest way to synchronise your computer’s clock”. Once installed it auto starts with Windows and can be launched from the ‘notification area’ (the bit by the clock). I recommend you select an NTP server from the list for the county you are in then hit OK, this is my config below;
Tropo & Aircraft Scatter (FT-8 to the rescue)
There are a large number of propagation types which enhance, or attenuate, VHF signals. Tropo scatter and aircraft scatter are 2 of the main propagation types which have made using FT-8 on 2m very popular, these propagation types can allow modest stations to complete FT-8 QSOs on paths over 600Km. These paths have always been there, but since FT-8 became popular on 2m, it has highlighted how common these paths actually are. Because the output power is constant when using FT-8, and the frequency is “fixed” and “stable”, it’s much easier to notice when FT-8 signals increase in strength, sometimes increasing so much that signals appear from nowhere for a few minutes then fade into nothing again – this is where experience and patience is needed. If the QSO path is via aircraft scatter, and the aircraft is moving at a tangent to the QSO path, then it’c common to see doppler on the FT-8 signal. Here in east Suffolk, near the flight paths for Luton and Stansted airports, it’s common to see a single station being decoded on 2, 3 or even 4 different AF frequencies due to all the aircraft reflections and the Doppler caused by their motion.
Modes and Bands
Each mode has a benefit over others, while SSB and HSCW are still used in Meteor Showers, like the Perseids, I’m going to stick with the MUM (Machine Celebrated Modes) JT6m, ISCAT & FSK441. Here are the main calling frequencies (centres of activity) for 6m, 4m and 2m for these MGM modes. 4m is a little variable as different countries have different allocations so you may need to check the cluster or log into a VHF chat page like ON4KST (registration needed) to see what activity is about, or make a sked to know where to have the QSO.
15 second periods.
If very busy or paths to USA then sometime 50.323 is also used.
MSK144 / FSK144
|2m||144.300||SSB / CW||SSB QSOs via MS is possible during the larger showers. HSCW is not used much anymore.|
30 second periods.
|30 second periods.|
* Not all countries with 4m are allowed to use the frequencies noted above, so note this when beaming in certain directions or making skeds.
Leaving your RX running on any of these modes with MSHV in RX (monitor only) mode, it’s highly likely you will see many reflections from a number of stations.
With my Icom IC-910 + 9 element beam YU7EF EF0209 @14m AGL and Icom IC-756pro3 + 5 element Duoband @12 AGL I’ve worked 30 stations in the past few days, here is a map showing some of these QSOs, and remember this is otherwise a closed flat bands where people say it’s not worth operating on!
Some Examples of Decodes
EI9E in IO43 on 6m JT6m using MSHV;
LW2WR and UT5DL on 2m FSK441 using MSHV, note LW2WR is RX’ing on 144.384 and UT5DL is RX’ing on 144.355;
DC2ECM in JO62 on 2m FSK441 using WSJT;
Points On TX
Once you are happy with RX and want to start TX’ing then Kevin, G0CHE, has a good description of the MS Message Procedure on his website, so I won’t reproduce it here, however some points I’d like to share are;
- Try and be mindful of those around you who might also be wanting to make MGM QSOs, and try and keep the MGM calling frequencies clear (don’t hold simplex QSOs on them if possible).
- Messages can also be called Macros, and start TX1 and got up to TX7 in MSHV (TX1 to TX6 in WSJT). You always send the next message to the one you received. So if you decode the message as shown in TX1 you would reply with TX2 and the other station after decoding your TX2 would reply with TX3…
- Some stations will TX with a CQ call on the calling frequency (as in the table above) but RX on another frequency. They do this so they don’t ‘hog’ the calling frequency. You can identify that as the CQ message which include the RX frequency in KHz, which on 2m would look something like “ CQ M1BXF 393 393 CQ M1BXF 393 393…”, this means I’m listening on 144.393 (running split) and if I hear an answer I’ll switch split off and carry on the full QSO on 144.393.
- If a station had lots of callers then they might not send 73’s. RX of RRRRR (TX4) is the official requirement for a QSO to be complete. So if I were working G3PYE and was sending “RRRR RRRR M1BXF…”, once G3PYE gets that message they can decide that instead of sending you 73 they reply to another station who had been calling them so you might decode something like “2M0ZFG G3PYE R26 R26…”. This means G3PYE decoded your RRRRR (and the QSO is complete) and G3PYE are now answering 2M0ZFG who they also heard calling. Now it’s likely 2M0ZFG was sending G3PYE a report which G3PYE decoded so they only need to reply with R26 (Roger + report), this means a busy DX station might only ever send TX1 (CQ) and TX3 (Roger Report) or even TX3 only if they are really busy.
MSHV vs. WSJT v10 (r5422)
After writing this article I thought I needed to run some more tests between MSHV and WSJT v10. With my IC-910 listening on 144.370 I started both WSJT and MSHV programs. The conclusion was MSHV has more accurate decodes, and in one instance meant I could have made a QSO when WSJT would not have allowed me to, I’m going to run many more comparisons as in truth although MSHV was better in these tests, there was in reality little difference between WSJT and MSHV was slight but this test was short. I also appreciate everyones millage (audio response out of their radio) may vary giving different results.
You will most likely have to click on the images to open a larger version in a new window to read the decoded text clearly.
In the first example the decodes are very similar, however the WSJT decode has an additional ‘1’ character on the message, what it does hint at is that in some instances such additional characters could lead to confusion, this might not be the best example to show my point due to the decent length of the decode, but assume the ping was half the length then the ‘1’ could cause confusion as to what the report could be; ‘26’ or ‘16’.
In this second example, WSJT decodes much less of the ping than MSHV did. The first line is the Auto decode, the additional lines are my attempt to Manually decode the message by clicking on the burst. Note that WSJT fails to decode is the QSY frequency, either Auto or Manual, where as MSHV decodes ‘365’ using either. Both WSJT and MSHV report the burst length as being 60ms long, so the fact here is WSJT just decoded less than MSHV, ultimately meaning that if I were using WSJT alone I would not have know what frequency to call BQI back on (ok the full callsign was not decoded, 9Y6BQI is not valid).
There is much more information on the web, much more than I want to include in this article. A few sources for additional information is John, G3WZT’s website and a video from a presentation given by Lyn GW8JLY I attended at the 2013 RSGB Convention on the topic “Meteor Scatter for Beginners”.
I hope some of you give this a go, and please get in touch if you have questions.