Feature Article
by Jon Grosjean, KB1SWW
Making a digital interface for your transceiver
I have a Yaesu FT-7900, but I suspect the following applies to most transceivers with the 6 pin DIN connector on the rear. All FM transceivers have a built-in limiter to prevent over-modulation. This is required by the FCC. On mine, it is a soft limiter so there is no hard clipping. It begins to limit at about 30mVrms input on the 1200bps input pin, pin 5. If you use the 9600bps input, it will limit at about 1Vrms. At limiting, the FM deviation is very close to 3.5kHz peak. Whatever interface you use, does not matter much, because the internal circuits will prevent over-modulation.
The FLDigi program produces modulated tones. Therefore the overall scheme with the transceiver is FM-FM modulation meaning the carrier is frequency modulated by frequency modulated signals. Therefore, if you produce a signal from the FLDigi program, and the frequency indication on the computer screen is not where it should be either at the center frequency or bandwidth, it is due to the software and its interaction with whatever sound card you are using. The deviation on the FLDigi screen is due to the sound card and has no relationship with the FM deviation of the carrier. If the transmitter is over-modulated, it will show up on the FLDigi screen on the receivers.
I measured the output of the sound card at maximum volume. It does not clip at maximum, so the best way to be sure you get the same results every time is to set it at maximum volume and adjust the interface to deal with it. In my case, the L channel of the PC put out .9vrms in Windows with the R channel disabled, and 1.26V with the R channel enabled (no explanation for this). With Linux, the output was 1.3Vrms. Using a cheap $15 USB StarTech sound card, the output was 1.15Vrms for both operating systems. If you are using Ubuntu Linux, and probably any other distribution, the cheap sound cards seem to work best because they use a generic sound IC for which there are drivers built in to Linux.
The microphone input on most sound cards has an AGC setting which can be enabled in the software. If it is not enabled, it will clip the input around 10-30mVrms. The accessory output on pin 12 (1200bps) or pin 4 (9600bps) is around 300-350 mVrms which is too high for most direct microphone inputs. If the AGC is enabled, the mic input will actually work without distortion with inputs from 3-100mVrms, so the attenuation on your interface box is not critical.
There is one pitfall I discovered with the StarTech USB sound card: The input jack is stereo, with the L and R inputs connected together, so if you use a mono phone plug, the input will be shorted to ground. Below is a schematic of my interface. The receiver output is attenuated to about 20mVrms for the sound card input. The transmit signal is rectified by a Shottky diode and drives a transistor to pull down the PTT line of the transmitter. If you still have some Germanium diodes around, they are great for this. It is important to make sure the sound card output is set to maximum, so it will produce a signal large enough to drive the transistor.
One point about using Ubuntu: The Pulse Audio program does not seem to work well with any sound card. I replaced it with the Alsa program and they all seem to work fine. When you start it, the output volume is at minimum and you must first move it to maximum. Also, make sure the Automatic Gain Control is set for the input. There is also a setting in Windows for this.
I actually use the 9600bps output with an RC lowpass filter for de-emphasis because it is not squelched, and, I think it is easier to see a signal in the noise.
updated April 6, 2011