Thursday, 1 December 2016

MSK144, and a lean November

It certainly has been a busy month at GM4FVM, but not on the bands. Antenna-wise there has been a lot of action, but radio activity has been less noticeable.

At last I have started to use the new mode MSK144, either in the excellent test version of WSJT-X 1.7, or in MSHV 1.28.

Perhaps you have a 6m transceiver, and maybe a half decent antenna. Well you can be on meteor scatter in a few minutes, working all over Europe with MSK144. So why not read on and try???

If you look at the WSJT official site you will find no mention of WSJT-X 1.7. The official site is promoting the last stable version of WSJT-X, 1.6.0, in versions for Windows, Mac and Linux, plus the source code. However, version 1.7.0 is out there in various test versions.

WSJT-X 1.7.0-rc2 in its various builds is a development on 1.6.0 but also includes a different suite of modes. Well, most people call them modes, but "protocols" may be a better term as they are all related. 1.6.0 offers JT4, JT9, JT65 WSPR, and Echo. Currently 1.7.0 offers JT4, JT9, JT65, WSPR, ISCAT, QRA64 and MSK144.

It is MSK144 which seems to have most appeal for me. There is no place in this latest trial version of WSJT-X for either FSK441 or JT6M. These two modes are in use mostly on 6m (JT6M) and 2m (FSK), with 4m dividing opinion as to which was best at that frequency. The key issues I saw were the FSK was not effective on longer bursts, which is what you tend to get on 6m, but JT6M's poorer error correction made it generate screen-fulls of garbage text. JT6M worked on Es, FSK struggled with those steady signals. Each had its own place on those specific bands.

MSK144 offers better performance in both respects. Using MSK, I find that it is equally happy with either long bursts or short pings, making meteor scatter easier. It seems to decode weaker signals than before, including inaudible ones. There are almost no garbage decodes. On VHF we seem to have settled on 15 second tx/rx segments, making QSOs, potentially, shorter. There is not a lot I do not like about MSK144.

If you get MSK144 in the WSJT-X suite there are several useful features. I have been using WSJT-X since version 1.3, and it is a joy to simply click on a callsign and reply without typing, entering reports or organising which segment to use etc. The 1.7.0 version bring this to MSK, and also adds "auto sequencing" whereby the software recognises the stage in the QSO, and steps itself through the contact without intervention from me. Well, I look on in amazement. I even transmitted to a station for 10 minutes and had given up, when the other station's reply arrived. The suite sent my RRR reply without me doing anything.

Now if you are like me, and read some of the amateur bulletin boards, you probably know how this is going to be received. "This is the thin edge of the wedge" ,"automatic QSOs, we might as well stay in bed" ... which is of course nonsense. You have to initiate the contact or noting happens. There are no automatic QSOs. It is real radio. Yawn. The level of debate on these sites is dreadful. I have watched a fine amateur being torn to shreds of admitting he went to bed and left his WSPR running. Yesterday I read that buying a Chinese screwdriver would bring about the end of liberty in the world. Here's the trick with these people - don't tell them what is new. They oppose anything new, but they never do any research or even reading. So, if we leave them to find out for themselves they will never bother us.

Anyway, yes, there is an issue about remaining in control. We need to show this thing who is master. To help, since v1.3.0 there has been a "watchdog" feature should you suddenly be called away to help chop down a tree and forget to stop transmitting CQ or anything else. My experience with v1.6.0 on HF suggests setting the watchdog to 10 minutes is good for calling CQ, if a bit long for other situations. For meteor scatter, 15 minutes might be best. After that time it simply stops transmitting and shows a red banner.

So this is all tantalising. A new suite with new modes and features. But you said it is not available, Jim? Yes, true. The official WSJT-X site does not even mention it. This is in part due to a complete faff over v1.6.0. As JT modes are open source then anyone can produce variations on them. It got crazy when a certain person put out his own version of v1.6.0 while it was still in development, and the developers got swamped with bug reports about something they had not issued and which was officially still being developed all the time. At one stage they asked other developers to back off a bit.

WSJT-X 1.7.0 is well enough advanced for many stations to be using it in development mode. It appears to be moving towards a "stable" version for the public as a whole. As I write (1/12/16), many stations are using MSK144 on 6 metres with WSJT-X, so it is getting a good test. However, I am not about to set myself up as a source for a test program. There is another way.

I have said before that I use Christo's (LZ2HV) suite "MSHV". Well, after a while during which the JT folks asked everybody to hold back on MSK144, Christo has had the go-head and has produced steadily improving versions of his suite.

Now, MSHV is a much simpler set-up than WSJT. The modes available include ones left off the JT version, such as FSK441 and JT6M, plus JTMS from some time ago. But it does have ISCAT and MSK144! So if you install the latest version of MSHV (v1.28) you are ready to go.

The MSHV website is here and it is simple to follow. Choose the installer version of MSHV if you want minimum hassle, whereas the ZIP versions allow you to retain a copy in a file more easily. Installation all goes the way these things do. After installation you set up your audio and PTT settings and you are ready to go. It is pretty light on processor power, but if you select "deep search" on the decode options it will use more PC capacity (but it will work better!). Tx/Rx is adjustable and pre-set to 30sec, but most stations using MSK are using 15second segments, whereas most using FSK and the other modes use 30 second segments.

MSHV is not "feature-laden" in the way that WSJT-X is. There is no watchdog, no automatic progression, no changing frequencies using CAT. But it works and it is available to all now.

There are a couple of issues with MSK. There are plenty of stations trying the new mode. I sort-of fear that the same thing will happen as took place after the "VHF features" were introduced to WSJT-X some time ago. Then a whole world of amateurs emerged onto VHF and made contacts on the new high-speed JT9 modes. Then they went back to using JT6M and FSK441. Let us hope that this time JT6M and FSK441 are quickly abandoned.

Also, the frequencies to use are a bit problematic. People seem to have decided that 50.280 and 70.280 are the places to go (though the band-plan says otherwise; 50.320-50.380 and 70.250). Why the opportunity was not taken to move to bandplan frequencies (say, 50.350 and 70.250) is a bit of a mystery to me. But there is no shaking the ability of amateurs to defy reasonable planning and just do whatever suits them. Who am I to disagree? I am just following the crowd, for lack of any alternative.
A 6m contact completed in 2 minutes ...
You can opt to report your presence on PSK reporter. This shows who is sending and receiving messages by band. It is useful to some extent, if you can drive it properly. More on that later.

I am sure that WSJT-X 1.7 will be widely available before long. However, in the meantime, if you can work any data mode you can download MSHV and use MSK144. Good luck.
Not much to report in November.

VK5ZK on 10m JT65 on 4 November set a nice tone, but there was nothing more like that. 10m and 40m WSPR were generally poor.

Meteor scatter
Meteor scatter is where most of the action was, FSK441 gave way to MSK144 here after 17 November. OZ1JXY, OH2BYJ and LA4LN, all regulars, were worked on 4m before that. Afterwards,
MSK accounted for DM2ECM, DF9OX, SM4GGC, SP8SN, G3XVR, OE5MPL, EA2ARD and GU8FBO on 6m, plus OZ1JXY and PA5Y on 4m. OK, no new countries and I have worked most of these stations before, but the contacts were quick and easy on MSK. Any I had not worked were new squares, showing how little attention I have paid to 6m before a better antenna arrived.

Eh? Not really expecting this. Very strange conditions on 25 November. There was that fraught contact with OZ1DJJ (not DZZ!) on 6m. I thought that was back scatter but he was far too steady for that. Then I heard Henning OZ1JXY (732km), very strong and steady, on 4m FSK. His signal was distorted and only decoding for some of the time, and this went on for about 10 minutes. I tried various modes before just calling him on SSB.
Very strange tropo on 4m
Henning reported my audio as Q4 at first, before QSB made a good QSO possible. Then I heard the OZ7IGY and PI7CIS beacons on 2m before a long period listening to the GB3PI repeater in Cambridge (446km). Tropo OK, but not as we know it. Really good on 6m, which is unusual, and even on 4m it is not usually that good. Atmospheric pressure was high, but there were few reports of other stations getting in on it.

There was short 4m opening on 10 November. I worked GM4OBD (IO97, a watery new one for me) and GM4ILS (IO87), but the distortion was terrible. It took OBD a lot of calling on CW to raise me. I was calling on SSB, and I thought  "well that is a funny noise" before it dawned on me that it was actually CW with no tone somewhere in amongst that huge aurora whoosh. ILS was trying to tell me something, but apart from the basic details I could make very little out. Later I heard the Faeroes beacon, OY6BEC, which is common on these occasions, and then it was over.

Thanks to MSK144 I have had a few contacts. Otherwise it would have been really depressing. But December is usually good for meteors, and maybe an aurora would brighten things up. And at least I now have a new mast, a better 6m antenna, a functioning power supply, a project to fix the old power supply, and I am ready for the new month.




  1. I agree about bandplan. I am starting to use 50.380 for MSK144. Why we need to use 50.2XX in the SSB part of the band for MGM defeats me. Let's start using the proper frequencies shall we?

  2. Hi Jim, Ive just started getting some gear together to come back to radio after about 20 years off air. I'm going to be on 6m predominantly as it holds most interest. MS is a favorite mode and Ive been reading your blogs with interest. I have got MSHV working in receive (bit of wire for antenna, awaiting LFA yagi).
    Question. When im on MSK144 mode in MSHV the rx report defaults to +00 etc like a FT8 report, not the 26 or 38 reports etc I am used to. I saw it was like this in your screenshot above. Is this method of reporting the standard. I see other MSHV modes have the traditional default duration and strength reports so it left me confused. Can you shed any light on this?
    David G0LBK IO93

  3. Hi David
    You are quite right about this. The previous meteor scatter modes released by WSJT did indeed use the 26 and 38 type of report. I seem to recall the 26 type report being used since meteor scatter came into common use with high speed CW.
    MSK is the first MS mode I have used which went over to the same system as WSPR and JT65, in other words decibels related to the noise level in a (theoretical) SSB-type filter width. So we now having something more like a signal strength report with MSK, or at least its signal to noise ratio.
    This move was not popular with some meteor scatter traditionalists. The strength of the pings is not usually as important as their spacing and duration.
    For me, I found the 26-type duration and strength report pretty limited. Either I got 26 (almost always) or 37 (rarely, meaning longer and stronger reflections). I never got report making much distinction between length and strength. If the reflections were weak and short it took ages I still got 26. It would have been better if people really used it accurately, but I just kept getting 26.
    Some people gave additional reports directly related to the number of reflections and their maximum length (like 3/500). That might have been "best practice", but others didn't bother. So it was fairly random.
    Yes, it is interesting to know how many and how strong the pings are, but with all the other WSJT modes using the dB response we now have MSK144 doing the same. You see prople giving the 3/500 type reports on cluster spots. Useful information I suppose.
    So there it is. The WSJT software has changed and our reports have changed too. Whether that is for the best or not I couldn't dare to comment.
    There may be some argument about MSK being so good we don't need to pay attention to the ping lengths any more but I would not hold with idea that at all.
    Good to have you back on the bands and I hope you get fully operational on 6m soon with that LFA.
    73 Jim

  4. Hi Jim,
    Thank you for the detailed reply. I'm not going mad after all!
    The MS system was always limited because of the mode. The signal report, 26, 27 etc.etc. was always based upon the first reception of a signal anyway and not always truly representative of the whole. QSL cards used to provide more info like "50 pings 10 bursts max 5 seconds S6".
    I've got the thing transmitting and all levels are adjusted. Just waiting on the antenna. I've set up MSHV and tried the WSJT but found it more awkward to configure; at this stage I prefer the MSHV, but that may change in time.
    Thanks again and keep the blogs coming.
    Hope to work you soon. 73 David