Doing the splits

At the end of last month I was testing a new version of WSJT-X for Joe, K1JT, under NDA. The program has now been released, so I can write about it.

WSJT-X 1.1 can decode both JT65 and JT9 at the same time

The major new feature is that WSJT-X 1.1 supports both JT65 and JT9. What is even more remarkable about it is that it can decode both modes at the same time! Enabling dual modes is optional, but if you use the feature and double-click on a station to reply to it the program will switch to the correct mode for the reply.

Another thing about this new version is that it can decode signals in a bandwidth 4kHz wide. If you have a transceiver that can receive such a wide bandwidth - Flex SDR radios can do this, as can the Kenwood TS2000 and my Elecraft K3 with FM filter installed - then you can take advantage of this capability. I didn't think my K3 could go that wide, but all that was needed was to run the K3 Utility and enable the FM filter in DATA mode.

If you're sharp then you will have thought of a snag. What happens if you reply to a station on the right hand side of the bandwidth? Receiving may be OK through the FM filter but on transmit the 2.7KHz SSB filter is used. The solution is clever: you keep the audio in the range 1 to 2kHz, engage SPLIT mode and set the transmit frequency to shift the signal up or down so that it matches the frequency of the station you are working.

I must admit that this perplexed me at first as I didn't understand the significance of operating in split mode. My audio tones (heard with the K3 monitor turned up a little) were often lower or higher than the station I was replying to, and I was afraid I was replying on the wrong frequency. In fact, I was: I had forgotten to switch SPLIT on!

To avoid mistakes in future I created a Windows shortcut using a little utility I wrote to send a CAT SPLIT ON command before starting WSJT-X 1.1 and equally importantly, set SPLIT OFF when I had finished a session with WSJT-X. It's just too much to expect me to remember to do this manually! I suggested to Joe that WSJT-X itself should send the split commands, but apparently it's not that simple when you have to cater for every transceiver under the sun. So if you are working JT9 or JT65 and  stations reply to you one or two kHz off-frequency don't be surprised.

I should make it clear: WSJT-X 1.1 doesn't decode both modes across 4kHz. It decodes JT65 in the lower half and JT9 in the upper half. You get to decide where the dividing line lies.

Joe K1JT thinks that JT9 users will move down a bit into the top end of the JT65 range, to enable people whose transceivers don't give them 4kHz bandwidth to take advantage o9f the ability to work dual modes. In fact, at the moment, the result seems to be to have enabled JT65 users to spread out over 3kHz or more as is clearly illustrated by my screenshot. Time will tell.

JT-Alert for WSJT-X

The eagerly awaited JT-Alert for WSJT-X has finally arrived! You can download it from the Ham Apps website.

This useful accessory will let you know if you have worked a station B4 or whether a station will fill a wanted band or mode slot. It sends spots of JT9 stations to the HamSpots website, providing a useful reverse beacon for the mode. It also logs contacts to  a few of the popular logging programs including MixW which happens to be the same log format used by KComm. Due to some nifty programming this new version of JT-Alert works with JT65-HF as well.

This new program couldn't come soon enough for me as I have worked just about everyone who is on JT9 at the moment and it was getting tiresome doing manual log lookups. Hopefully this new program will attract some new participants to this amazing mode.

W(h)ither JT65A?

I was looking at all the stuff in my ham software folder and thought about deleting JT65-HF. I no longer see the point in continuing to use JT65A when an alternative exists which uses the same messages and takes the same amount of time but is more sensitive and much narrower. The fact that this alternative has come along at the same time as W6CQZ, the author of JT65-HF, has stated that he can no longer develop it seems to me very fortuitous.

Okay, there are still a lot more stations to work with JT65A but that is surely only temporary. Like Tim, G4VXE I can see activity switching from JT65A to JT9-1 over time and JT65A activity on HF withering away from lack of use. JT65A has been a lot of fun but it can also be frustrating when you can't find space to operate within it's 2kHz or when a very strong signal stops the software from decoding anything. JT9-1 has people using too much power too, but the decoder seems more robust and a lot more stations can concurrently use the same bandwidth. It just seems an all round better solution.

I expressed this opinion in one of the Yahoo! groups this morning and several people replied to disagree with me. As one person put it "if I am half way to WAS (worked all states) using JT65A then I am not going to stop using it until I have finished it."

To me, I can see no reason to continue using a mode once a better alternative exists, which is why I have no interest at all in using RTTY, for which there are several superior alternatives. The idea of making contacts just to put a tick in a box in a matrix ordered by band, mode and prefix/country/state makes me quite depressed. I will use JT9 until something even better comes along, after which I will lose interest in it just as I have lost interest in JT65A.

Ghostly signals

I have been puzzled and mildly irritated by the 'ghost' traces that appear at plus and minus 100Hz when I receive a very strong signal with WSJT-X.  Torben OZ1TMK wrote to me after we had a JT9-1 QSO to ask if his signal had been OK. He had received requests from a couple of local stations to reduce power because he was 'causing harmonics'. I hadn't noticed anything wrong with Torben's signal but it hadn't been strong enough. It looked to me as if it was the same effect I and a few other JT9-1 users had observed when very strong signals were received, so I decided to investigate.

I had a theory that the +/-100Hz spurious outputs ( +/-120Hz observed in the USA) were caused by ripple modulating the transmitted carrier. I used my general purpose signal generator, otherwise known as the FT-817ND, to transmit a low power carrier (CW key down) into an unscreened dummy load (Elecraft DL-1). I repeated this with the transceiver powered from my bench power supply and then on its internal batteries with the power cable removed. The results in the WSJT-X spectrum display window are shown below.

K3 RX, FT-817 TX on mains supply.

K3 RX, FT-817 on battery power

As can be seen, there are traces at +/-100Hz and at 100Hz intervals on both signals, but the ghosts seem a bit stronger on the signal when the TX is powered from the main supply.

I recalled an issue a few years ago when someone sending CW using their K3 had reports of spurious signals +/- the sidetone frequency. This turned out to be audio modulation of the synthesizer by the sound of the sidetone from the K3 speaker. Elecraft provided a fix in the form of a stiffener for the synthesizer board. My K3 is an old one and does not have this modification. You can see that the synthesizer is affected by physical vibration looking at the trace produced when I rapped on the K3 case.

K3 RX showing the effect of vibration (knock on the case)

My K3 sits on a shelf next door to a heavy linear power supply. Could slight vibration of the mains transformer be modulating the receiver's local oscillator so as to create weak sidebands at +/- double the mains frequency?

To answer that question I repeated the tests using my Elecraft K2 as a receiver, feeding the headphone output at low level into the cheap USB audio dongle I use for computer sound. You can see the result below.

K2 on RX, FT-817 on mains power.

K2 RX, FT-817 on battery power

You can see that the sidebands are much reduced when the signal is produced by a transmitter running on battery power. In fact, some weak +/- 50Hz sidebands are present - possibly the effect of 50Hz AC hum on the un-isolated cable used to connect the K2 headphone output to the sound card.

I'm not sure what to make of all this. It does appear that the 'harmonics' - which are really sidebands - that accompany a strong JT9-1 signal are caused mainly by AC ripple modulating the transmitted carrier, but that hum on the receive side can produce a decodable signal as well. The WSJT-X software is extremely sensitive and can detect these components even if they are 30dB or more below the fundamental carrier.

I would appreciate hearing of other theories or tests carried out to explain this phenomenon. It seems to be that this issue is going to be almost unavoidable when mains-powered equipment is used to generate signals that are decoded by very sensitive software.

EX8MT Kyrgyzstan

This afternoon JT9-1 gave me my first contact into Kyrgyzstan, with EX8MT.

Looks a bit like the English Lake District!

Far East DX

After yesterday's disappointing session on 15m PSK31 I tried JT9-1 again. The new JT mode showed its superiority by giving me two contacts with Japan. I nearly worked more as one JA replied to my CQ but disappeared after I sent his report. I was also spotted on HamSpots by two other JAs whom I have never seen on my screen. Another nice contact was with HS0ZBS. This is not the first time I have worked Thailand, but all my contacts with that country have been on 15m.

In another direction I worked PJ2MI - a familiar callsign. Checking my log we first worked in 2003 using PSK63 - quite an achievement as I would have been using my K2 and 4W output. More recently I have worked Jose using PSK31 and JT65A. On 20m I worked K8CZC who from the look of his eQSL would have been using low power (QRP).

New Yahoo! group for JT9 mode

It was inevitable this would happen: someone has started a Yahoo! group for the JT9 mode. Pity they couldn't have made it a Google group like the one for JT65A but I guess more hams have Yahoo! accounts than Google accounts.

There was already a fairly quiet WSJT group but this will keep the discussions about using JT9 on HF separate from the rather esoteric MS and EME stuff.

Richer pickings on PSK

I only managed 4 QSOs in abour as many hours using JT9-1 this morning. I worked LU8EX whom I recognized having made a JT65A contact with him in the past. Hopefully more operators will make the switch from JT65A to JT9-1. At the moment it feels like I've worked everybody. I was spotted many times by VK3AMA even when I was running 5 watts. Pity there is no-one else in Oz using the mode yet.

I switched to PSK31 in the afternoon and my QSO rate immediately improved. A nice catch was Luc PR8EP whose QSL card I picture here.

Another good one was Eric HS0ZJK in Phetchaburi, Thailand. He is only the second Thai station I have worked, and both were on 15m.

JT9-1 with 1W

I see that some bloggers have complained that HF propagation today was not too good. My 5W JT9-1 signal on 20m was spotted several times at reasonable strength by VK3AMA so this afternoon I thought I would see how far I could get with 1 watt.

Reception of JT9-1 signal from G4ILO using 1W to attic dipole

Well, my 1 watt signal didn't make it as far as VK but the map above shows how far it did get in little over an hour's operating. I called CQ continually except when my QRPp was answered by calls from RV9WF, R9WJ and S51AY. I'm pretty pleased with that. This JT9 is shaping up to be one heck of a digimode!

Spotted in VK

If you get the impression that I'm keen on K1JT's new JT9-1 mode, you'd be right. This morning my signal (25W to an attic dipole) was spotted by Laurie VK3AKA in Australia - the longest distance I've been received using JT9-1 yet. A QSO must surely follow.

My signals down under were -18dB. Yes, it seems Laurie is testing some new add-on for WSJT-X that spots to Hamspots with additional info. Exciting stuff! JT9-1 is turning out to be the WSPR QSO mode that many hoped for.

WSJT-X update

Stay out of the shack for a few days and up pops another new version of WSJT-X.

Version 0.8 revision 3112 sports a number of improvements including CAT control (see the frequency box.)

At first I thought CAT control wasn't working because if I changed my K3 frequency with the VFO or the band up/down buttons the changes were not reflected by the software. Turns out that's how it is supposed to work. You can set the rig frequency by means of a drop-down list above the level control but it doesn't work in the other direction.

Just spotted there is a newer revision already! Download it from the usual location.

A rant about RTTY

I just worked Iraq for the first time. YI1RZ on 20m using JT9-1 mode.

YI1RZ

What was remarkable about this QSO is that I worked YI1RZ despite the presence of heavy RTTY QRM. My Elecraft K3 has great filtering but it can do nothing about a RTTY signal that is straddling the QSO frequency.

Why, in this age of DSP, do people persist in using this antiquated mode? Dating from the 1930s and the age of mechanical teleprinters and analog modems, RTTY is by any definition an outmoded mode. It requires far too much power and occupies far too much bandwidth for the data transmission rate (45 baud.) PSK31 has existed for more than a decade and is a far more efficient mode. If PSK31 is too slow then there is PSK63 which is faster than RTTY and yet still manages to occupy less bandwidth. Not to mention the plethora of other modes such as MFSK and Olivia that have been invented in the last 10 years offering far greater reliability than RTTY and, like PSK, the ability to use the entire character set not just capital letters, numbers and a few punctuation symbols. And which don't print up garbage because a shift character wasn't decoded.

It feels good to get that off my chest! I'd better put my asbestos suit on! It is interesting to note that as I have been typing this WSJT-X has been pulling JT9-1 out of the air in the teeth of RTTY interference so great that you cannot even see the JT9-1 signal traces. JT9 rocks! I doubt if the RTTY operators even know the JT9-1 signals are there.

New page for JT9 modes

JT9 just got even better. Thanks to Laurie, VK3AMA, JT9 has a new page to show current activity or general chat at hamspots.net.

As far as I can tell, the spots are generated from the spots to PSK Reporter, which the latest WSJT-X program does automatically if you tick the appropriate box. So you don't have to do anything extra for your spots to show up on this page. This is a great facility that I am sure will increase the growing popularity of the JT9 mode.

New WSJT-X update

Hot news. I've just had an email from Joe Taylor K1JT informing me that a new version of WSJT-X, v0.6 r3045, has just been made available for download from the WSJT home page.

This is a must-have update which now includes spotting of JT9 signals decoded to the PSK Reporter website. (Tip: This needs to be enabled in Settings.)

Still to come are rig control - which I will hazard a guess will use hamlib - and integration with VK3AMA's JT-Alert. Hopefully this will encourage current users of JT65 on HF to try the new JT9-1 mode which is 2dB more sensitive than JT65A and occupies less than 10% of the bandwidth.

Thanks Joe for giving us these great programs.

JT on JT9

I had an all-day session on the JT modes today. Actually, almost all of today's contacts were on JT65A. I did listen for JT9-1 signals a few times but most of the time I only saw stations I had already worked before.

I put out a CQ call on 15m JT9-1 and Joe Taylor K1JT came back to me. I've decoded Joe's WSPR signals numerous times and I've exchanged several emails with him but I had never had a radio contact with him until now, so that was a nice surprise.

Then it was back to JT65A. The contrast in activity was extreme. There were so many stations active I couldn't find a space to call CQ, so I had to wait and pounce on new stations that called.

I must say that when using JT9 I miss the infrastructure that has been built up around the JT65A mode - the reverse beacons, the auto-spotting to PSK Reporter so you can see how far your signals have got, the JT-Alerts when you decode someone you've worked before. I especially need the B4 alerts. My memory is so bad I can't remember the calls of stations I've worked before so I have often called people who I have worked only a couple of days earlier. I guess that due to the lack of new stations they probably don't mind too much!

Ghosts on JT9-1

I spent a couple of hours working JT9-1 on 20m this afternoon. I only messed up once when I forgot where I was in the sequence. I could do with the software showing the outgoing message in the log as well.

I think word of the new mode is spreading slowly as even on a Sunday there were not many new stations to work. N1ISA(FN41), WB2LMV(FN21) and KC2RDC(FN14) were the only DX worked.

One Italian station was a record +24dB during one call. I reduced power to 1 watt but I still received a +00 report from him. I don't know how much power he was using but it is a bit unfair to accuse a station of using too much power when even QRP signals are strong. WSJT-X seems to hold up better than JT65-HF in the presence of a very strong signal, as even in the presence of signals well above 0dB I was still getting decodes.

One thing I have noticed when receiving strong signals is that I get duplicate 'ghost' decodes at 100Hz intervals on either side of the main trace. Look at the four decodes of IK3SCB at 1701. There must be some 100Hz modulation of the signal, but I don't know if the cause is at my end or his.

WSJT-X update

A couple of days ago I had an email from Joe, K1JT, author of the WSPR and WSJT software. He had read my post about my first JT9-1 QSO in which I said that I missed the JT65-HF user interface. Joe pointed out that WSJT-X is in a very early stage of program development, and user input will surely help to define its future evolution. He asked what features of the JT65-HF GUI I found desirable.

I replied with what I thought were the key points that made JT65-HF easier to use. The result is a new version of WSJT-X which I have just tried. One change is that the horizontal 'panadapter' display scale now matches the waterfall when the user has set FFT Bins/Pixel greater than 1.

However, the real big change is that double-clicking on a decode line now generates a set of messages addressed to the second callsign on the line, regardless of where you double-click. It also sets the Tx and Rx frequencies to that of the decoded transmission and selects the first message in the sequence. This is a big time and error-saver in the few seconds you have between receiving a call and having to reply. You still have to set Auto to ON to enable the transmitter and select the next message in the sequence after the first has been received. Perhaps it's a matter of personal preference but I don't think it is a bad thing for the user to take control of this rather than have the program try to work out the appropriate reply. In other words, double-click on a decode when it is a CQ call or a reply to your CQ. Use the Tx n buttons to select the next message in the sequence as you progress through the QSO.

Try this latest version of WSJT-X. I think you'll find it a big improvement. Now all we need is for Laurie VK3AMA to come up with a version of JT-Alert that adds logging and 'worked before' detection and there will be no reason not to switch to this much narrower JT mode.

My first JT9-1 QSO

PC4T Paul's blog post about working Tasmania with 5 watts gave me the spur to try the new JT9-1 mode, so I installed the WSJT-X software. The user interface is quite a bit different to the older WSJT programs but most of the same controls are there. I never really figured out how to use WSJT, much preferring the simpler interface of Joe W6CQZ's JT65-HF application.

Working SM5CS using JT9-1 mode

My experience with JT65-HF stood me in good stead as I was familiar with the sequence of exchanges, but I missed the JT65-HF user interface, its ability to decode all the signals in a swath of spectrum, and the alerts and built-in logging of it's companion JT-Alert application.

My first QSO, also using 5 watts, was with SM5CS - not as impressive as Tasmania but sufficient to satisfy myself that I knew how to drive the program. I'm puzzled by the panoramic display though: the two peaks of the spectrum analyzer display don't match up with the two traces shown on the waterfall.

I made the changes to KComm to allow me to log this new mode. I seem to have an increasing number of modes that I can log but not upload to eQSL.cc because the ADIF specification doesn't yet include them, though JT9 is already there.