Friday, 23 December 2016

WSJT-X 1.7.0

Thanks to many of you who notified me that WSJT-X 1.7.0 has now been officially launched on the WSJT website here.

This brings to an end the testing phase, when the trail versions were only available on Sourceforge.

Time to see what WSJT-X can do. I will report here on some issues I have noted so far. Some aspects like QRA64, which is an experimental mode for Earth-Moon-Earth communications, have not been tried by me yet so I will have to leave them for later.

1) Configurations
As I use different rigs for different purposes, I find this setting very useful. I have stored configurations for HF (FT450), 50MHz (IC7100), 70MHz (IC7300) and 144MHz (IC7100). Thus I can simply select the configuration and WSJT selects the rig control, sets the frequency and connects to the correct audio stream.

To make some sense of this from a ergonomic angle, I keep the HF activity to the top left of the PC display screen, 50 and 144MHz to the bottom left and 70MHz to the top right (the logbook is on the bottom right and the rotator control centre top). If I do not do this I get mightily confused (and the rigs are arranged in similar fashion on the bench).

WSJT-X 1.7.0 has a built-in solution for this, as not only does selecting the right configuration select the rig and its frequency, it also places the display in the correct part of the screen depending on which configuration I select. Now that is definitely clever. Add to that the fact that it can store the power levels by band and so forth, and that becomes a very powerful rig control tool.

2) PSK Reporter
PSK Reporter has been an option on WSJT-X as far back as v1.3.0 (and possibly further back). Whereas WSPR has its own dedicated reporting system in the excellent site, the other modes do not have anything similar which would allow simple "reverse beacon" operations. You can set this option and then you have to make sense of PSK Reporter, which is a bit clunky in operation.  It shows stations using various modes, including of course PSK. You can establish which mode is in use by reading the roll-over text boxes of the stations on the map.

This allows you to call CQ and then check PSK Reporter to see who has heard you. "But don't stations who hear your CQ reply to you?" I hear you say. Not likely, in my experience. I now know that I am heard by three or four times the number of stations who call me back.

And, yes of course, I pay them the same compliment by ignoring them too.

For several versions of WSJT-X we have had WSPR included.
WSPR is a low power beacon mode. In its WSJT-X guise the waterfall is much wider than the 200hz band in which WSPR stations live, so you can easily see stations who happen to be slightly mistuned.
The only possible issue I have with this version of WSPR is that the power slider allows you to carry over your power settings from other modes. You can opt to choose to carry over the settings per band, but not per mode.

WSPR is typically run using powers of less than 5 watts, often very much less than 5 watts. So a preset calibrated power attenuator for WSPR like that one that existed in WSPR 2.21 etc would be very useful, for me anyway.

This is not a big problem so long as you remember to check it, and WSPR operators are usually very careful about setting their power output carefully.

4) JT9 and JT65
As with previous versions of WSJT-X since around v.1.5.0, it is possible to run both JT9 and JT65 simultaneously, on receive at least. You have to choose which one your transmit on. This is especially useful with SDRs which allow for wide rx filter settings. The JT9 frequencies tend to be 2kHz above the JT65 ones, so with a (say) 4kHz filter you can listen to both.
You can see here that the higher rx reports, which were for JT65 stations, have a # mark, whereas the later JT9 ones are marked with a @. You need not worry too much, as if you want to reply to any of them, you just click on the station, and WSJT selects tx on either JT9 or JT65 as appropriate, moves to the correct frequency on the waterfall, inserts the callsign into the generated messages, and puts the signal report in automatically too.

If you did click on the image above to look more closely, you will see that I worked DL1FAM at a signal strength of -25dB. Now this is a figure taken against a reference figure for noise in a standard receiver using an SSB filter and it can be taken with a pinch of salt. Yes, the signal was weak, but not 25dB below my noise floor. I could also hear stations on the speaker at -10dB, but I could not hear that station at -25dB. So it was certainly weak. JT9 gives about a 2dB improvement over JT65. I would say that JT65 is a bit better than a good morse operator. So -25dB was certainly a weak signal. Not perfectly received though, as I missed one rx segment, but then QSB is always with us.

Far and away the majority of my 6m QSOs are made using JT9 or JT65, and pretty well all of my HF ones. JT65 was originally designed for use on Earth-Moon-Earth paths, but it is now firmly rooted on the HF and VHF bands for F and e-layer communication as well.

5) MSK144
This new mode offers advantages over JT6M and FSK. These include faster data rates and much better error correction. The general convention to use 15 second tx/rx segments does fox me a bit at times, but I am getting used to it. Unlike the previous modes in use, MSK144 seems at home on both the longer meteor bursts of 6m and the shorter ones on 2m.

I have reported before how I stupidly managed to get the configuration wrong somewhere along the line of setting up. I am now using RX 1500hz, F Tol 100, T/R 15 seconds. I have decode set to "deep". Deep will use more processor power than fast or normal.
The % figure in the "receiving" box indicates processor power used.

If you tick "Auto Seq" you allow the suite to complete your QSO once you initiate a contact. It will incorporate the reports and progress down through the QSO format as it receives the appropriate replies. In the absence of replies it just sends the same message (but updates the reports!). Note that the "TX Watchdog" (also a carry-over from previous versions) will eventually stop you transmitting if you fall asleep while this is working automatically.

This works well enough, save for a few niggles. If you receive more calls from a station before their report arrives, the suite keeps updating your report based on the last signal received. I have once sent five different reports before I received mine. Let us hope that the other station did not get confused. Also, some stations have sent me out-of-sequence messages, such as sending their first report as, say R-02 rather than -02, which sets the Auto Seq off course. But provided you stay in control this setting has the advantage of speeding up contacts and reducing use of the spectrum It certainly helps you get the best out of a long meteor burst.

The automatic setting for split frequency (entering a working frequency while calling CQ on a common frequency), is something I have not yet got working with the IC-7100. Nor have I used the "sh" box for shorthand messages, which I feel are not necessary for my type of operation.

6) Summary
WSJT-X 1.7.0 is another step forward.

Much has been said about MSK144, a new mode which seems by all accounts to render existing meteor scatter modes outdated. Apart from the new modes, this marks another step to incorporating all the "JT" modes into the WSJT-X wrapper. For me for instance, the "configurations" setting is a real plus. In the past people struggled to get on the air with JT6M or FSK441 using WSJT9 or 10, which did not interface well with modern rigs. Now they can use MSK144 with WSJT-X, and benefit from CAT control.

The WSJT-X suite brings us to another milestone  - the multi-band suite. Now, in theory anyway, you only need one suite to work HF, Es, E-M-E, meteor scatter, using modes such as JT9, JT65, QRA64 and MSK144. You can switch between rigs using configurations, use certain modes on certain bands or with certain rigs, and generally never actually touch a VFO or a mode control.

Ham Radio Deluxe and FLdigi may have been good HF suites, but WSJT-X is becoming spectrum-wide.

I have said it before, but it deserves saying again. If you can set up your rig audio and CAT as for, say Ham Radio Deluxe or any similar data mode, you have got exactly the right setup to run modes covering many aspects of radio activity today. The days of having several different pieces of software to run WSPR, HF data modes and meteor scatter are long gone.

Unless of course you are mad enough to run all your rigs simultaneously as I do.

But them I am a bit crazy.




  1. This comment has been removed by the author.

  2. Split CQ operation is well implemented.

    IC-7300, in WSJT-X preferences choose "fake it' for the split mode. That works best in this case, but experiment if your rig is different,

    To call CQ tune your Rig, or set it from WSJT-X to your RX frequency say 50.350
    Choose TX CQ Frequency in the program and tick the box to right of that.
    Note CQ message TX6 becomes CQ 350 GM4FVM IO85.
    Once a partner calls you the rig will automatically change over to TX and RX on 350.
    Even neater if you reply to a station calling say CQ 380 GM4JJJ IO86 who you hear on another frequency, then your transmitter will automatically shift to transceive on 380 for you. No need to touch rig, use split or any other witchcraft.

    Excuse brevity of comment, the first time I typed a more detailed instruction, but it was lost when I tried to preview it.

    Enjoy, and Merry Christmas.

  3. OK, David. Thanks.
    I have tried this again amny times but could not get it work with either IC-7300 nor IC-7100. Using "rig" or "fake it" might help, but the freqeuncy I am going split too is the same as one I am calling on, so it gets me nowhere.
    With the TxCQ spinner setting at, say, 385, and the box ticked, the frequency in the generated message does not change from 370 even if I try to regenerate std messages.
    So, of course, the VFOs keep changing back to the 370, or 280, or whatever they are associated with in the mode settings.
    This did not happen in the past as far as I can recall, so I ran 1.7.0 rc3 (7364) to see, but it follows the same path.
    As you say, it depends on the rig, but both of them seem to be doing the same thing.
    I think I will need to play around with this a bit more. Perhaps I am doing something wrong, which is obvious to everybody else and not obvious to me. If so, anyone who spots it please let me know.

  4. Hi Jim,

    thanks for the article!

    With respect to working random MS from a calling frequency using rig control. You set your rig to the frequency you want to QSO on and set the "Tx CQ" spin box to the calling frequency where you will call CQ. WSJT-X will send CQ messages on the "Tx CQ" frequency but QSY to the original rig frequency if you get a reply. We changed the sense of this for the v1.7.0 release as it works much better. Sorry for any confusion from prior release candidate versions.


  5. Hi again Jim,

    your use of configurations in WSJT-X is a good one but if you have rigs running simultaneously there is an other way. You can run multiple instances of WSJT-X simultaneously, all that is required is each one is started with a unique "--rig-name=" command line option. It looks like you are running on MS Windows, you can copy the WSJT-X icon and then right click, go to Properties->Shortcut->Target and add --rig-name=some-name-of-your-choice to the end of the target command.

    You can also combine configurations and multiple instances, for example you might switch between WSPR and JT65+JT9 on an HF rig instance and between 6m and 2m MSK144 settings on a VHF rig.