Hello Janne, > > You might try the "SOUNDCARD TEST MODE" of Linrad. Send a single > > carrier into your RF hardware and look at what the audio card > > sends into the digital hardware. (Works as a two-channel oscilloscope) > > You should see two sinewaves that are of the same amplitude and > > shifted by 90 degrees. If you see something else, maybe just a single > > sinewave, something is wrong and it should be fairly easy to trace it. > > > > You may download Linrad.exe that should run under any Windows system > > that has properly installed drivers for the soundcard. > > http://www.sm5bsz.com/linuxdsp/linrad.htm > I installed the software but couldn't get any signal on the screen, > just noise. But ok, I just have to study the program :) These ASIO > soundcards are driving me grazy, especially E-MU, the software is just > too complicated. Btw, this was my first touch to Linrad. You have done > some awesome work Leif! ASIO??? Linrad can not use ASIO drivers, the soundcard must be installed with drivers that conform to the Microsoft standard. At least this is what I have been told - I do not use Microsoft myself so I really do not know. (I prefer Linux.) If you were presented with a choice to select the soundcard during setup, and if you see anything at all, the audio and computer side should be ok.
> Just a hypothetical question, what if my local oscillator signal > isn't pure? I use 74AC74 as a /4 and am feeding it with a simple > oscillator, which is basically a inverter (I think 7404) crystal > oscillator (14.4MHz). Can this be the cause for my IQ balance error? Maybe if oscillations are not governed by the X-tal at all. Such an oscillator may oscillate with the X-tal as a capacitor at some frequency far from the X-tal frequency. You may verify by listening with an ordinary radio at 14.4 MHz and then at 3.6 MHz for the fourth sub-harmonic that the 74AC74 should produce. If you get good tones, all is OK. Then try to route the loudspeaker output of the radio into the soundcard, one channel at a time, to check check the channels individually. When using the Linrad "SOUNDCARD TEST MODE" you should place a test tone close to the LO frequency to start with because Linrad just shows the pixels without drawing lines between them and if you display a high frequency your eyes will not connect the points properly and you might think that a curve that varies greatly from pixel to pixel is just noise. If you sample at 48 kHz, it is a good idea to test with an audio frequency of 2.4 kHz or less to have 20 pixels or more for each period of the sine-wave. 73 Leif / SM5BSZ
