Hi Ian,
Thank you for finding some precious time to answer to me.
Regarding oscillators I had tested and my subjective findings:
NDK NZ2520SD (29.4912 MHz) - Very musical, a lot of bass slam and on the same time a lot of information in bass area. This Japanese clock is highly sensitive to power supply. Better to be powered by batteries than ADP151. I have to repeat the tests with my own designed power supply based on a supercapacitor.
CCHD957 (24.576 MHz) - It is more tolerant to power supply. Sound different than NDK, not good or bad, just different but it is not so resolute in bass area.
Generic Kyocera XO (29.49 Mhz) - Resolution is lost with this XO even powered by batteries.
Other generic XOs (29.49 Mhz) - Some of them are pretty good but as long as I can not trace the manufacturer I can not rely on them
All tests were made by changing oscillators and power supply in my own designed DAC based on DSD1792 and AD1896 ASRC. As source I had used SDtrans384 powered by Eneloop batteries followed by a CLC filter.
I've never tested SI570 but comparing the specs with NDK and CCHD957 I've find it inferior.
Kind regards,
Mihai
Thank you for finding some precious time to answer to me.
Regarding oscillators I had tested and my subjective findings:
NDK NZ2520SD (29.4912 MHz) - Very musical, a lot of bass slam and on the same time a lot of information in bass area. This Japanese clock is highly sensitive to power supply. Better to be powered by batteries than ADP151. I have to repeat the tests with my own designed power supply based on a supercapacitor.
CCHD957 (24.576 MHz) - It is more tolerant to power supply. Sound different than NDK, not good or bad, just different but it is not so resolute in bass area.
Generic Kyocera XO (29.49 Mhz) - Resolution is lost with this XO even powered by batteries.
Other generic XOs (29.49 Mhz) - Some of them are pretty good but as long as I can not trace the manufacturer I can not rely on them
All tests were made by changing oscillators and power supply in my own designed DAC based on DSD1792 and AD1896 ASRC. As source I had used SDtrans384 powered by Eneloop batteries followed by a CLC filter.
I've never tested SI570 but comparing the specs with NDK and CCHD957 I've find it inferior.
Kind regards,
Mihai
Good ear, Mihai. Did you try sync mode with AD1896? I have an ESS9018, it comes with similar internal ASRC architecture than the AD1896. I found it sounds quite different at async and sync mode. It’s kind of different style, although both of the modes sound wonderful. Actually ASRC couldn’t eliminate jitter, it only shift the spectrum of jitter. So, I look ASRC upon as a digital filter. Where your MCLK came from? The finial sonic performance of an ASRC based system will not only decided by the local clock, but also combined with factors of MCLK. That’s why I’m suggesting trying the sync mode. I believe you will get some new discovery from those XOs.
Thank you for sharing your experience. I did too, both async and sync mode. But I wouldn’t want to tell how the Si570 programmable XO was compared with CCHD957, NZ2520 and TentLabs, you know the reasonJ. But I have to say Si570 is much much better than that normal Kyocera XO, as well as the regular ones from CW and Fox. Chance to have a 22.5760 one, but don’t know if it’s as same as yours.
Regards,
Ian
Status update: GB II will start getting delivery by this weekend.
Loop testing was done. It’s much faster than I thought. All of the finished PCBs are 100% tested and bit perfect confirmed.
The only thing I have to mention is, some of the 11.2896MHz verification oscillators(on the single XO clock board) have bended pins. They just look ugly but function wise no any problem. I should plug them into the sockets by myself at the last stage rather than ask the SMTA supplier. Sorry about that. Please let me know if you couldn't put up with those ugly XOs
.
All of GBII items are ready for delivery now. I’m busy processing the orders, will start getting delivery this weekend.
Have a good night.
Ian
Loop testing was done. It’s much faster than I thought. All of the finished PCBs are 100% tested and bit perfect confirmed.
The only thing I have to mention is, some of the 11.2896MHz verification oscillators(on the single XO clock board) have bended pins. They just look ugly but function wise no any problem. I should plug them into the sockets by myself at the last stage rather than ask the SMTA supplier. Sorry about that. Please let me know if you couldn't put up with those ugly XOs
.All of GBII items are ready for delivery now. I’m busy processing the orders, will start getting delivery this weekend.
Have a good night.
Ian
Servus Ian,
The MCLK in my DAC is generated locally, not recovered from SPDIFF interface, and is common between DSD1792 and ASRC .
How do you suggest to put the DAC in sync mode with the transport, getting the MCLK from transport and bypassing DAC clock and, also, bypassing the SPDIFF receiver and connecting directly I2S signals to DAC?
Kind regards,
Mihai
Ian,
Great work...thank you for your diligence. Will be very happy to get this completed as the last step in my project.
Gary
Sorry for the confusion. Yes I mean shipped or posted. They will be delivered soon.
Ian
Ian,
Great work...thank you for your diligence. Will be very happy to get this completed as the last step in my project.
Gary
Thanks Gary, you will get the KIT in two days.
Good luck.
Ian
they can all be supplied from the one PSU 6v @ ~600mA for the lot. depending on the dac the only one that may benefit from being powered separately would be the clock board. not sure exactly the draw and it will vary depending on what speed and type clocks you are using. probably around the 100-150ma though for 45.x/49.x