[christianw.]

Bezüglich des DACs, da kommt jetzt die "KIMI RÄIKKÖNEN" Variante. (Leave me alone, I know what I'm doing)

Zitat: Originally Posted by doctormord View Post Well, every single resistor got its own parasitic capacity, no matter how the layout is routed. (Even soldering joints are relevant) Beside this, your switching is done via MOSFETs? Those gates are capacitiv coupled via the bulk, so there are more than a few pF. Clean plots? Easy explanation. At 384kHz the 470pF capacitors have a capacitive resistance of ~880Ohms, so 384kHz wont be possible with full output signal scale. (Would guess, even 20kHz will be damped) Conversation error induced harmonics are just shorted this way. A 384kHz frequency plot would show this.

Zitat:The switching by the mosfets inside the LVC595's is not an issue driving the resistors, more a power supply decoupling issue.... And of course the resistors have parasitic capacitance, like between any two signals, but those are sub pF. Lets just say I'm not worried about the parasitic capacitances anywhere in my DAC design. Although I'm not a RF designer, I design Ghz computer boards in my day job so I have some experience there.... 384 Khz sampling is more for show anyway, afaik there are no 384 Khz music files available, I could as well do 768 Khz sampling.... I have have one more analog bessel low pass filter in the output drivers, have decided on a system -3db at 200 Khz, 12 db per octave passive filter, which in my opinion is perfectly fine, there isn't any music up there.... But feel free to remove those capacitors