21.09.2005, 08:55 AM
SSasen,
delayed answer is not problem, you just can say what you know about it right here. Did you try 8939 with symmetrical supply actually? Or can you measure the impedance between 1&2 pins? Thank you!
Selfoscillating around 8939 will occur without external comparator (BTW, why not only BJT pair, instead whole comparator chip to the 8939 input?), if you will compensate bias from negative supply rail e.g. through same res from the Vdda, or if you'll use extra integrator for keeping almost zero DC. I don't measure output noise exactly so far, because my ears not hear such noise amount anyway, by the other words, i have added 20db loop gain else by the extra integrator (9 BJT else), so THD(.0025% -3db), IMD(.01% -10db) and noise well enough for me. UcD not in use, just hysteresis modulator. You can find Philips reference PCB layout and schematic (for symmetrical supply) here http://www.alcom.ru/intros/ivx.rar
delayed answer is not problem, you just can say what you know about it right here. Did you try 8939 with symmetrical supply actually? Or can you measure the impedance between 1&2 pins? Thank you!
Selfoscillating around 8939 will occur without external comparator (BTW, why not only BJT pair, instead whole comparator chip to the 8939 input?), if you will compensate bias from negative supply rail e.g. through same res from the Vdda, or if you'll use extra integrator for keeping almost zero DC. I don't measure output noise exactly so far, because my ears not hear such noise amount anyway, by the other words, i have added 20db loop gain else by the extra integrator (9 BJT else), so THD(.0025% -3db), IMD(.01% -10db) and noise well enough for me. UcD not in use, just hysteresis modulator. You can find Philips reference PCB layout and schematic (for symmetrical supply) here http://www.alcom.ru/intros/ivx.rar