For audio clocking purposes long-term stability and accuracy is irrelevant, as almost any crystal oscillator will be good enough, yet that is what TCXOs are designed for. Short-term phase noise/jitter is the issue; that depends mainly on the quality of the crystal and the circuit design, not the temperature compensation. The only benefit of a high quality expensive TCXO is that there is a chance that the manufacturer has used a high quality crystal.
I use a highly modified Musical Fidelity X-DAC V3 via coax SPDIF from an Asus Xonar D2X soundcard. For best results the "filter" pin capacitors with either the DIR1703E used there, or the DIR9001 which replaced it , should be replaced by close tolerance polyprops or other high quality types.Some time back I replaced the standard 24.576mHz oscillator with a 1PPM Vanguard TCXO .It sounded much the same initially, then after several tracks you would become aware of the SQ having lifted to a higher level. Recently I replaced the 1PPM type with a
.3PPM Vanguard TCXO.There was another lift in SQ. About 2 weeks ago I fitted a
5V Zener referenced opamp supply, a little similar to a "PFM Flea", to supply this TCXO instead of the existing 7805 which was already supplied by a type of "Shunt Regulator" (JLH). There was a further quite noticable improvement with 24/192 material like 'Norah Jones-Come Away With Me" and "The Eagles -Hotel California"despite being downsampled to 24/96 to suit the X-DAC V3, then it upsampling again to 24/192. High res material was much more noticably improved than RB CD after the PSU improvement, although it benefitted too.
These improvements were NOT placebo, as they were verified by another DIY Audio member who then installed my recovered 1PPM TCXO in his similarly modified X-DAC V3 which we had been comparing mine with.
Alex
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.