I have had exactly one FT-8 mode contact on my FT-817, so I am now declaring myself an expert. I am sure there are aspects of this that I haven’t discovered yet, but what I have works and will get a field test on my next SOTA outing. Before I forget how it all works, I thought it best to make some notes.
My basic set up involves a laptop, a SignaLink USB interface (model 6PM), and my trust old FT-817. I went with the SignaLink as a convenient way to get sound in and out of the radio. The unit comes with a USB cable to connect to the computer and a data cable that plugs into the DIN connector on the back of the FT-817. The SignaLink provides line isolation for the audio signals and solves the problem of needing adapter cables and perhaps an attenuation pad between the radio and the computer. Because the audio is routed through the data connection, the volume knob on the radio can be twisted up and down to listen at a comfortable level without affecting the signal level that the computer sees.
I followed the instructions provided by SignaLink for configuring jumpers so that the right signal is on the right wire leading to the DIN connector; no problem there. Windows10 did the right things when I connected the USB cable, so no problem with driver set up. I had previously configured the computer to not make any system sounds, so I left it that way. Within the sound set up, I turned the volume output for the SignaLink to 100%, as instructed by SignaLink.
I connected the radio to my antenna, tuned to 14.074 Mhz and heard a wall of FT8 signals. I launched WSJT-X to make sure that sound was making it through to the program, but instead of seeing the signals on the waterfall as expected, up came a dialog box because of course the program couldn’t find Elecraft K3 that it normally sees on its serial port.
Initially, I went into the File>Settings>Radio and tweaked the radio setting to FT-817, but that is also wrong unless you really have a CAT control cable for your FT-817. I don’t. I’ve used FT8 without CAT control before and it generally works well, even if it isn’t as pretty as having everything automated. For me, the right way to configure the radio connection is “None”. The only other setting on the page was PTT method, which I set to VOX.
The other WSJT-X setting that had to be changed was the audio tab, to specify the SignaLink as the device for sound input and output. Once I did that, my waterfall lit up and decodes started to appear.
With the SignaLink receive audio set to 50%, the audio was a bit hot, so I went into the computer’s sound settings and dialed down the SignaLink device input sensitivity to about 25%. I then dialed the rig to a quiet frequency and adjusted the SignaLink such that the background noise was registering at -30 dB on the WSJT level meter.
It wasn’t too hard to get sound moving in the other direction. When I clicked on “tune” I expected the SignaLink to generate some kind of sound into the radio. However, when I plugged earbuds into the monitor jack on the back of the SignaLink, I heard nothing. I played with the sound settings in WSJT-X and the problem turned out to be that the setting “mono” did not generate sound, but it worked fine when set to left channel or both. When I played around with the output level setting in WSJT, I found a level that would trigger the PTT built into the SignaLink. When that happened, the rig would key, as expected.
Note that the rig’s VOX should not be set since what turns transmit on and off is actually the signal from the SignaLink via the data cable.
Although the rig keyed, when I set the meter to PWR, I saw no power out, and confirmed that there was no modulation in SSB mode. After scratching my head for a moment, it occurred to me that I needed to set the radio to digital mode in order to pipe audio in via the rear data port rather than the microphone port. I went into rig setting #36 and set the digital mode to PSK31-U. That mode applies no audio shift, so if the computer modulates a 1khz tone, it will go out 1khz above the dial setting, just as you would want.
When I switched to DIG mode, the waterfall contracted to 500Hz bandwidth, which makes sense since I have a narrow filter installed for CW (and in principle, it would also be helpful for RTTY or if I were in an FT4 contest). I wanted to see the whole audio passband, so I deactivated the filter and the waterfall re-expanded to the expected 2300Hz bandwidth of the SSB filter.
I then played around with the combination of WSJT output level (about 75%), SignaLink transmit level (about 25%), and the DIG MIC setting (menu item #35), which was set to 50%. I adjusted mostly the SignaLink transmit level until I stopped deflecting the meter when set to ALC mode. As far as I know, the rig is not applying any other sort of compression, equalization, etc., so the signal should be as pure as it can be as it leaves the rig.
With all of this seemingly working, I switched the rig over to the dipole on my roof and sent a couple CQs on a clear spot on the waterfall. After a call or two, I got a reply — thanks Sergey, UT2IF. The automated exchange ran its course.
A few moments after I made my first contact I realized that the radio had been set to the lowest power. Sites that received my 500 mW signal lit up in the next few minutes and are shown above.
So, the take homes: 1) follow SignaLink instructions on configuring the SignaLink; 2) set the radio tab in WSJT-X to “None” with PTT method to VOX; 3) set the audio tab to use the SignaLink as the audio device and make sure that sound comes out of the SignaLink when you hit “tune” in WSJT-X; 4) set the radio mode to DIG, set the DIG mode to PSK31-U; 5) play with windows sound output level, the output level in WSJT-X, the SignaLink transmit level and the FT-817’s DIG MIC level such that the radio’s ALC never kicks in on peaks. That should essentially be it.
Now, when it comes time to actually use this in the field, there’s one more twist: time synch. Many of the peaks in Cyprus have line of sight to a cell tower, so I should be able to tether the computer to my phone and get an accurate internet time. For the more remote locations, the computer has NMEATime2 installed and can take its time reference from GPS satellites.
So, now to wait for a bit of good weather and see how all this can go hilariously wrong when I actually try to use it in the field. More to come.
Addendum 1: My PC laptop is not of recent vintage, also it’s not very light. For peaks that are a longer trek, I’d prefer to carry my Macbook Air. The SignaLink manual made it sound like it would be difficult to set up. It turns out, that it was a breeze. I installed the most recent WSJT-X for Mac per instructions including the readme that comes in the .dmg, and then just duplicated my configuration (>File>Settings on the PC, WSJT>Preferences on the Mac). I then set the sound output for the SignaLink device to 100% and Mono. For time synch, I forced an update from an ntp time server with the terminal command “sudo sntp -sS time.apple.com”.
At that point, signals were decoding with no problem on the waterfall. I pressed “tune” and adjusted the power slider in WSJT to just under the point where I got ALC deflection on the FT-817. Again, a couple calls at 500mW resulted in spots across Europe on PSK Reporter. I upped the power to 5W, made a test call and worked stations in the Netherlands and Belgium.
So, I now have two options for operating FT8 portable. Since both computers are a few years old, it would make sense to do a battery life test before this weekend to get a sense of how long each will last with the screen moderately bright (sunlit peaks) and also powering the SignaLink.
Addendum 2: The night before taking this whole contraption out for field testing, I noticed that I receive signal would intermittently drop way down. I tapped the SignaLink and it came back — just the sort of intermittent fault you wouldn’t want to have to deal with on a mountain top. The problem turned out to be that one of the configuration jumpers provided with the SignaLink was a bit loose in its socket. I think the provided wires are 24Ga and they are somewhat brittle. I replaced these with my usual 22Ga jumpers, which seemed to fit the sockets much more securely. Generally, I think the engineering and construction on the SignaLink is very good, but this does seem to be a weak point.
Addendum 3: I did try this in the field and posted the results on this blog.