El 06/03/2010 14:53, Rein A escribió: > > Hello Jose, > > I always set the sound card volume, the modulation, that when changing the > volume setting, the output of the transmitter will follow in a linear fashion. > This is very important in particular for WSPR and WSPR-QSO modes. > > 73 Rein W6SZ
I do likewise. My homebrew interface has no variable adjustments at all, and I do all the settings in the computer. I use a professional (even when small) 600:10000 transformer, backwards, so it acts as an attenuator, towards the radio mic input. I use a small pot core ferrite transformer as 1:1 ratio isolator, loaded with 1200 ohms and 2.2 nF to get the least overshoot in the square wave edges. Even when I am going to send mostly sinewaves (band pass, 300 to 2700 Hz audio), it gives a measure of received bandpass flatness. That is the radio to PC channel. I noticed a slight hiss/harshness in the highs reduction in the PC speakers when the transformer stopped ringing. I listen thru my 2.1 speaker set, which sounds better. I use a 4N26 optoisolator with a red LED in series (visual PTT indicator) shunted by a reverse connected 1N4007 that was at hand, to protect the LED and optocoupler, and a series resistor I believe is a 2.2 K resistor (do not remember clearly now). All the paths are isolated, but the PC and radio PSU are connected to ground, a couple of rods and a big old truck radiator buried in the garden.My metal desk is also tied to ground, which allows me to work with static sensitive components with total confidence. I have done eventual envelope checks with my oscilloscope (a -40 dB tap in the SWR probe), to make sure there is no envelope clipping at normal levels. I also check routinely the tx level when I change bands. I built a PEP (peak holding) SWR indicator, and always look for a slight decay in the output while setting the soundcard output. It assures me there is no clipping in the chain from the soundcard to the antenna. I usually do that with the TUNE button of MultiPSK. 73, Jose, CO2JA
