Saturday 29 December 2018

DATA Socket

I'd been using an FT-847 with a data modes interface to my PC, completely isolated with transformers in both transmit and receive paths and a mechanical relay to switch the PTT. It used the 3.5mm jack on the rear of the FT-847, which is what you must use for transmitting on SSB (the usual 6-pin mini DIN is for FM transmit only).
Everything worked OK on data modes like FT-8 and WSPR, but I noticed that when making a recording of voice, there was something wrong with the frequency response. At first, I thought that this output might be bypassing the de-empasis circuit on FM, but when looking at the frequency response, even on SSB it rolled off below 1KHz.
The transformers I used in the interface were from a 'ground loop isolator' designed for car audio systems, and when tested between the output and input of the PC, showed a response within 2dB down to below 100Hz.
Eventually I thought about the transformers and their impedance. 600 ohms is a common value for them, maybe there was a mismatch between that and the output impedance of the radio.
Checking the schematic for the FT-847, the final op-amp in the 'RX1200' circuit goes to a 100nF capacitor, then this goes through separate 4K7 resistors to both the 3.5mm and DIN sockets.This would explain why there was such a big cut in low frequencies.
At the worst case, driving a short circuit, the 4K7 and 100nF would make a high pass filter with a frequency of about 340Hz. The specification of the data output says a 10K impedance, and with a 10K impedance (which would be in series with the 4K7), the frequency would be around 100Hz. That's fine for any voice or data modes, even 340Hz would be usable.
But the transformer will also create a R-L high pass filter, with its impedance increasing with frequency (DC resistance is about 100 ohms so increasing from that), against the 4K7 resistor. This output is not really suitable for driving an isolating transformer if you want anything like a flat frequency response.
The resistor and capacitor could be replaced with different values (higher capacitance, lower resistance).
I had a look a schematics for other radios made by Yaesu, that had a 6-pin data socket. Although they all used similar op-amps to drive the RX1200 output (NJM2902), 3 radios had 3 different combinations of resistor and capacitor in series.
FT-847 - 4K7 / 100nF
FT-817 - 1K / 1uF
FT-8900 - 220R / 1uF
That's a 20:1 range of output impedances, although every time this is specified as for use with 10K input impedance on whatever equipment is connected. If you're making a direct connection from the radio to the Line In of a PC or amplifier, it probably won't matter as the input impedance will be quite high. But the circuits for most home built and commercial data mode interfaces have isolating transformers, and will affect the frequency response.

Thursday 22 February 2018

VHF FT8

After reading on Twitter that some people were trying the very popular FT8 data mode on 144MHz, I had a go. I don't have anything like a VHF DX station at home, with just a Comet GP15 vertical on the chimney, fed with a bit too much coax.
The frequency for 2m FT8 is 144.174.
I started off using my new SDRPlay receiver and found that I was able to decode a few stations, outside my usual VHF range (G4KUX in IO94 and GM4FVM in IO85).
After connecting up a transceiver, I was able to make some 2-way contacts at this distance, and was even getting reception reports of my signal via PSK Reporter from up to around 300km. This makes the 2m band a bit more interesting, and because data modes software like WSJT-X automatically sends reports to PSK Reporter, even if I'm away from the radio, it's still worth leaving it on all day just in case someone calls CQ.
At the moment, there aren't really any 'lift' conditions on 144MHz, at least not which you would hear if listening to FM or even SSB, and most of the long distance propagation is from reflections off aircraft. Reflections off fast moving objects like aircraft shift the frequency from the doppler effect, and it can be over 100Hz. With FT8, the bandwidth is very narrow, so if there are multiple paths, the same signal can appear on 2 different frequencies. See below for some FT8 activity where there are multiple decodes.

On some days I've received 10 different stations on 144MHz FT8.