I have not been on the air much over the summer. Too many other things going on. I've plodded forward towards DXCC and now have 95 counters worked on 40m, with 76 confirmed. On occaision, I still get poor audio reports however. Not the bad reports that I was getting when the PA stage was biased into class C, but gentle comments indicating that my audio sounds like it has too much base. This is my homebrew T2 transmitter based on the KK7B designs. It's part of the Gateway 40 transceiver.
To investigate I fed audio tones into the microphone input from an HP function generator. I set the level low, about 5mV, to make sure that there was no distortion caused through overdriving a stage. The graph below shows the output from the transmitter for single tones across the passband
The graph above shows the output of the final 5W PA. I also measured the output of the first RF stage and it shows the same curve. High frequency tones are attenuated.
I checked the output of the audio processeor next, measuring its output (at the input to the diplexer) and measuring the output of the diplexer as it enters the mixers. I measured both I and Q channels, which showed the same curves, although they had a slight ampliture difference. The graphs are shown below:
The outout of the audio processor seems fine. It attenuates audio below 500Hz and is flat across the rest of the audio band. I know that it also drops off rapidly after 3kHz, because of the filter in the microphone amplifier stage.
The output of the diplexer starts to show the trouble. While not as pronounced as the RF output, the graph has a similar shape.
I was not really sure what is going on here. The diplexers should be flat across the audio band. So I posted a question on the EMRFD Yahoo group and got a response from Rick Cambell KK7B. He had spotted this same issue with the T2, which is due to an impedance mismatch between the modulator and the diplexers. I added 47 ohm resistors in series with the I and Q channel outputs of the modulator, at his suggestion. I reran the rest and got the result below:
This is much flatter! It looks like a good response curve. Now to test it on the air!
Enter Comments Here:
Copyright 2001-2021 Chris Thompson
Send me an email