This weekend I have managed to get the FreeDV program working successfully and made a couple of QSOs on HF.

FreeDV has received a fair degree of publicity recently and is a GUI application for Windows, Linux and MacOS (other platforms are in development) that allows any single side band radio to be used for low bit rate digital voice communication.

I was primarily interested in it for possible use on VHF. It is hoped that the recently introduced 146-147MHz allocation available in the UK to full licence holders (via a Notice of Variation) will be used for digital mode experimentation. I have been constructing a data mode interface for use on the old TR9000 since it can operate on these frequencies but I decided to try the FT857D on HF to get an understanding of the program.

With FreeDV speech is compressed down to 1600 bit/s then modulated onto a 1.25 kHz wide 16QPSK signal which is sent to the mic input of a SSB radio. On receive, the signal is received by the SSB radio, then demodulated and decoded by FreeDV. The appeal being communications should be readable down to 2 dB S/N, and long-distance contacts should be possible using QRP power.

There is a lot of information. videos and guides available on the FreeDV website but the basics are straightforward. The computer requires two sound cards to handle the audio to and from the radio and the speaker/headphones and microphone. One method is to use a USB ‘gamer’ headset but since I already had a spare PC headset with 3.5mm plugs I opted for a cheap plug in USB soundcard (as detailed below)

Once the audio devices and PTT control are installed and configured within FreeDV it is a matter of monitoring a frequency (in USB) and hopefully you can send and receive.

So far I have only managed two QSOs with Trygve (LA6UOA) and Fabrice (F4FPG) on 14.236MHz with both the audio was clearly understandable but did suffer from the odd breakup and stutter. This was probably due to the laptop being used rather than an issue with the signal as the FreeDV needs a reasonably powerful machine with a decent amount of memory.

The software is very picky, several times the audio device settings have been forgotten or it has come up with a cryptic error message:

../src/msw/bitmap.cpp(846):assert “image.IsOk()” failed in
CreateFromImage(): invalid image

This is a known issue with the Windows version and the solution is to manually remove all of the FreeDV settings from the registry. The full details of how to edit the registry are in the GOTCHA section of the FreeDV site. Editing the registry is not recommended to the inexperienced person so is not a great solution and when starting the program backup after this fix requires reconfiguration of callsign, sound card and PTT settings again.

Despite this setback it is an excellent and exciting new mode. Both Trygve (LA6UOA) and myself were first timers when we had our QSO and we were like excited school children once we were able to converse successfully. Fabrice (F4FPG) was an old hand at it and I thank him for allowing me to briefly interrupt a QSO he was having with another UK operator who I could not hear.

I would recommend using K7VE’s QSO Finder website to see who is monitoring and on what frequency as well as passing hints in real time.

Datamode Interface built

This weekend I finally got around to sorting out my digital/data mode interface for the Yaesu FT-857D.

To transmit and receive digital/data modes you need to connect the radio audio in/out to the computers sound card in/out, the computer then runs the necessary software to encode/decode the signals. I want to try out WSPR, PSK, JT65 and some SSTV for starters I have spent too long just receiving and decoding…

There are a number of inexpensive commercial interfaces available, but many of them use the same basic design originally intended for eQSO/Echolink operation. I nearly succumbed but I had built an eQSO interface many years ago when using PMR446 and had most of the parts to build another.

I nearly took the easy route and got a commercial one since connecting up some home built circuitry to a £20 hand held is slightly less daunting than plugging it into an expensive rig!  My original interface has been modified and reused over the years and was a bit of a mess, but being brave I decided I could tidy it up and I couldn’t really damage anything if I took my time… actually the truth was I discovered I didn’t have the necessary optocoupler IC so couldn’t build a new one just yet…

A simple internet search for digital/data mode interfaces will bring up a great deal of information, schematics and ideas for home brew solutions. The basics can be found here for example.

The simplest form of interface is just a simple direct lead with the transmitter operating in VOX mode. However levels can be a problem as the line/speaker output from a computer can be too high for a transmitters microphone input. Also connecting a radio to a computer directly can lead to problems with ground loops and interference.

The computer can be made to control the Push-To-Talk (PTT) on the transmitter using a serial port with the software controlling one of the handshake lines (RTS/DTR) Some data mode software support CAT to allow control of the PTT as well as tuning the transmitter, but the serial port method is more universal.

The preferred interface, and the one I had built isolates the computer from the radio by using two audio transformers and an optocoupler. There is no direct connection between the two devices so keeps interference to a minimum.

I could have used the microphone and speaker output on the radio, but the FT-857D has a convenient data connector on the back. This is a 6-pin DIN socket as used by older (PS/2) computer keyboards/mice. Note the diagram shown in the FT-857D manual (as below) is the view as you look at the socket.

FT-857D Data connector as in the manual

The connector has two “data” outputs but they are simply fixed level audio outputs from the receiver. The one of interest for most modes is the 1200 baud output (the 9600 baud output is more akin to a discriminator tap and is only of use for FM packet) There is a data-in (TX-audio) and a PTT control line.

Like many people I initially thought I could cut a lead off a mouse/keyboard and repurpose it, however I discovered most only use four wires and they don’t use the necessary pins! You might be lucky especially with older keyboards or alternatively if you have an old keyboard extender cable they usually have all six wires present. Alternatively the plugs are readily available from the likes of CPC/Farnell.

I had a hunt around in a junk box and located a suitable keyboard extender cable. I chopped off the useless end and metered out the pins to identify the appropriate wires. Remember when looking at the plug the pins are swapped left-to-right compared to the diagram which is the socket view.

Well here is the insides of the interface.. and as you can see I completely failed to tidy it up! Not my best work, but I did put it in a new box and I did tape up all those unused wires!

The messy internals of the interface

One annoying issue I had was the audio from the computer wasn’t getting to the radio, it worked and scoped out okay when out the box, but in the box it stopped working. I suspected a bad solder joint and redid them all, but still showing the same intermittent issue. I did notice flexing the board slightly out the box cause the fault and soon located a track on the vero-board which appeared to have a hairline crack, I couldn’t see anything obvious but a wire soldered along it cured the issue.

It looks better with the lid on..

I soon had it connected up and fired up WSPR

Computer, radio and interface

It was straight forward setting up WSPR to use a combination of CAT for tuning and the RTS PTT control and soon had some encouraging results, in fact these are some of the spots of my 5W signal on 10m/20m and 30m, I was grinning from ear to ear!

10M Spots

20M Spots

30M Spots

I had simply set up the FT-857D for basic USB transmission, however it does have a dedicated DIG mode, the manual refers to setting it up for RTTY/PSK, I very briefly had a go with PSK31 and Digipan and was successfully decoding signals and put out a few CQ calls (again on 5w) but got no immediate response. However checking PSK Reporter later it seems I had been heard by VC3S, OH1FOG and ES1JA maximum distance was around 3260 miles

Looking forward to spending some more time experimenting with the data modes.