Interesting thread, I just came across with it. I made Latch signal reclocking in a modified Marantz CD-74 (TDA1540Ps replaced to a TDA1541A). I think Bit Clock is not critical here, since it is not I2S.
I also made DEM reclocking to 12 x fs (529.2 kHz), feeding Q and Qneg from a 74HCT74 to pins 16 and 17 (Cosc removed).
Someone mentioned excessive dissipation of the TDA1541A earlier in this thread. I glued two copper heatsinks on top of it, available in PC parts shops as memory cooler. The heatsinks are connected to pin 14 with a piece of wire.
I experimented with pin 4 and found no difference in sound, be it connected to pin 2, to GND, or to pin 9 of SAA7210 (the SAA7220P/B is bypassed).
Another observation is with the DEM filter capacitors. I added 22uF tantalum to MSB and MSB-1 (pin 12, 13, 18, 19). But I realized that the linearity is badly affected, and it is very sensitive on mechanical stress. It is clearly shown on an oscilloscope using the test method described here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=113620
I still use the SAA7220 as clock oscillator. The SAA7210 data sheet says it also can be used as oscillator (between pin 18 and 19). Would it be better than keeping the SAA7220 fro clock oscillator? I know the best solution would be a separate low jitter clock.
I also made DEM reclocking to 12 x fs (529.2 kHz), feeding Q and Qneg from a 74HCT74 to pins 16 and 17 (Cosc removed).
Someone mentioned excessive dissipation of the TDA1541A earlier in this thread. I glued two copper heatsinks on top of it, available in PC parts shops as memory cooler. The heatsinks are connected to pin 14 with a piece of wire.
I experimented with pin 4 and found no difference in sound, be it connected to pin 2, to GND, or to pin 9 of SAA7210 (the SAA7220P/B is bypassed).
Another observation is with the DEM filter capacitors. I added 22uF tantalum to MSB and MSB-1 (pin 12, 13, 18, 19). But I realized that the linearity is badly affected, and it is very sensitive on mechanical stress. It is clearly shown on an oscilloscope using the test method described here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=113620
I still use the SAA7220 as clock oscillator. The SAA7210 data sheet says it also can be used as oscillator (between pin 18 and 19). Would it be better than keeping the SAA7220 fro clock oscillator? I know the best solution would be a separate low jitter clock.
Actually, when I had a board (slight redesign) run, I found that I no longer needed the extra delay, so it is Q/ of the 74HC74 only now. I removed the inverter chip... not sure why as I was unable to run comparisons between the home-etched board and the commercially run board.Fin said:
- Status
- Not open for further replies.