@JensH:
Hi Jens. This is the datasheet I've found on NDK's webpages. There's also an application note, however, I reckon it's rather superficial ... The data Bunpei gives is beyond either of these documents, though ...
@marce: Can I ask you how it would be to mount the clock (NZ2520SD) directly on the pins of the ADC (and in the process ensuring good connections to ground etc.)? Esstech's ES9112 (to mention one) have a pin spacing that would allow this (1.27 mm) - could mean virtually no issues with reflections (I assume?), very low inductance and close to zero capacitance ... There might be an issue with crosstalk (I hope I use the right word) but there's a DGND pin between the "clock in" pin and the other datapins.
Also, if a low thickness PCB board with a ground plane just below the clk pin is used then any potential clock noise should not be able to go far ... ?
P.S.: Haven't yet read your links so might be the answer is in one of those!
Greetings,
Jesper
Hi Jens. This is the datasheet I've found on NDK's webpages. There's also an application note, however, I reckon it's rather superficial ... The data Bunpei gives is beyond either of these documents, though ...
@marce: Can I ask you how it would be to mount the clock (NZ2520SD) directly on the pins of the ADC (and in the process ensuring good connections to ground etc.)? Esstech's ES9112 (to mention one) have a pin spacing that would allow this (1.27 mm) - could mean virtually no issues with reflections (I assume?), very low inductance and close to zero capacitance ... There might be an issue with crosstalk (I hope I use the right word) but there's a DGND pin between the "clock in" pin and the other datapins.
Also, if a low thickness PCB board with a ground plane just below the clk pin is used then any potential clock noise should not be able to go far ... ?
P.S.: Haven't yet read your links so might be the answer is in one of those!
Greetings,
Jesper
Last edited:
Hi Gentlevoice, I would imagine that would work, basically the closer the clock connection the better, the signal arrives before any reflections etc can disturb it. Leaving room for a possible series resistor is recommended when using oscillators, just in case the output does need it, with direct xtals they are very rarely required. With buffers they are always required as these have quite hard current drive, for driving longer lines, so can often be problematic to get right.
Guys, fascinating thread and thanks for the information. Did anything come of the potential group buy? What were the results of testing? Is it worth putting an order in for some of the NDKs?
The reason I ask is that I am experimenting with upgrading the clocks (and their power supplies) on a PC motherboard to see what the effect is on the quality of audio output. I should stress that this is subjective, but changing 2 of the onboard crystals for a couple of Fox XO's from Farnell does seem to reap benefits. I'm currently powering the XOs with lifepos so I assume that the clean power is contributing in some way (and potentially is the sole contributor to the improvement - as it is impossible to power the legacy crystals externally).
I have currently replaced 2 clocks (the RTC @ 32.xK and the southbridge 25MHz) with header pins and have succesfully booted when either of the clocks is replace - but not both! When both clocks are replaced...no boot.
Apologies if this is straying too far OT but I have posted more details about why I think this is happening here:
http://www.diyaudio.com/forums/digi...t-xo-soundcard-how-deal-voltage-mismatch.html
...and I wondered whether you guys might take a look and see if you have any suggestions about how to reduce the output of the clocks and hopefully to get the MB to boot with both clocks replaced.
Thanks all,
Crom
The reason I ask is that I am experimenting with upgrading the clocks (and their power supplies) on a PC motherboard to see what the effect is on the quality of audio output. I should stress that this is subjective, but changing 2 of the onboard crystals for a couple of Fox XO's from Farnell does seem to reap benefits. I'm currently powering the XOs with lifepos so I assume that the clean power is contributing in some way (and potentially is the sole contributor to the improvement - as it is impossible to power the legacy crystals externally).
I have currently replaced 2 clocks (the RTC @ 32.xK and the southbridge 25MHz) with header pins and have succesfully booted when either of the clocks is replace - but not both! When both clocks are replaced...no boot.
Apologies if this is straying too far OT but I have posted more details about why I think this is happening here:
http://www.diyaudio.com/forums/digi...t-xo-soundcard-how-deal-voltage-mismatch.html
...and I wondered whether you guys might take a look and see if you have any suggestions about how to reduce the output of the clocks and hopefully to get the MB to boot with both clocks replaced.
Thanks all,
Crom
Hi Crom,
I can't answer your questions about listening except that I have tried to place an NDK NZ2520SD just in front of the DAC (following the Amanero Combo 384) and the sound was - in short - (much) more tonally nuanced & integrated, much more "quiet" in the positive sense and the perspective was also improved.
I case you are interested in trying them out they are available in smaller quantities here:
NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK
Best regards,
Jesper
I can't answer your questions about listening except that I have tried to place an NDK NZ2520SD just in front of the DAC (following the Amanero Combo 384) and the sound was - in short - (much) more tonally nuanced & integrated, much more "quiet" in the positive sense and the perspective was also improved.
I case you are interested in trying them out they are available in smaller quantities here:
NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK
Best regards,
Jesper
NDKs
I have used the NDK oscillators in a project. They replaced Fox Espressos, and the sound was quite obviously better from a subjective standpoint.
I also had the phase noise measured by a colleague, and it is important to note that some did not meet spec, and some were better. So if you know anyone who has the gear necessary to measure close in phase noise, it might be worth it to do so... This same colleague also has noted that he finds similar production variances with other well known low phase noise parts, like the Crystek CCHD Series.
Gentlevoice, what clocks were you using previously?
I have used the NDK oscillators in a project. They replaced Fox Espressos, and the sound was quite obviously better from a subjective standpoint.
I also had the phase noise measured by a colleague, and it is important to note that some did not meet spec, and some were better. So if you know anyone who has the gear necessary to measure close in phase noise, it might be worth it to do so... This same colleague also has noted that he finds similar production variances with other well known low phase noise parts, like the Crystek CCHD Series.
Gentlevoice, what clocks were you using previously?
Hi Barrows,
Interesting information you provide ... I look forward to using the NDK oscillators in my DAC hopefully soon to come ... With respect to your question I have only compared the NDK oscillator (49.152 MHz) with the oscillators present on the Amanero board (through the Amanero circuitry) so I have not made a direct comparison. However, placing the NDK oscillator close to my test-DAC made a quite audible difference. I suppose that optimising the PSU for the oscillators may also make a difference as well as optimising the PCB layout (I currently use PTFE 35um double-sided, 0.5 mm thickness).
Best regards,
Jesper
Interesting information you provide ... I look forward to using the NDK oscillators in my DAC hopefully soon to come ... With respect to your question I have only compared the NDK oscillator (49.152 MHz) with the oscillators present on the Amanero board (through the Amanero circuitry) so I have not made a direct comparison. However, placing the NDK oscillator close to my test-DAC made a quite audible difference. I suppose that optimising the PSU for the oscillators may also make a difference as well as optimising the PCB layout (I currently use PTFE 35um double-sided, 0.5 mm thickness).
Best regards,
Jesper
Hi All,
I'd like to add a few comments on my experiences using the NZ2520SD in my new (& approaching completion) DAC. The setup is as follows:
Amanero Combo384 -> Acko Reclocker -> own make new NZ2520SD clock board that feeds the DAC and the Amanero (45 & 49 MHz; more comprehensive PSU filtering) -> PCM1794 NOS (DDDAC inspired but different PCB & semi-active output stage; no capacitor). Dual mono altogether.
Impressions on this clock in the new DAC have been "correlated" with a previous PCM1794 version where I briefly had a chance to listen to that DAC with an NZ2520SD compared with the clock coming directly from the Amanero.
My impressions on using the NZ2520SDs: There's really no comparison to other DACs I've listened to in terms of tonal nuance, ambience reproduction, instrument stability, layering of instruments and the ability to discern even small nuances tonally, space-wise, or what could be called a general ability to nuance or "detail". The new DAC with the NZ2520SDs is also very rhythmically stringent: No lagging, sluggishness - just speedy, on time (this, however, I reckon may also be a DAC quality more so than having to do with the NZ2520SD).
To my ears the differences are not subtle ... I tend to compare the NZ2520SD differences with how I've experienced battery powering a full audio system may change the sound - yet battery power cannot "fill in the blanks in the music that are not there" due to a less stable clock ...
The reproduction with the NZ2520SD can also be "sweeter", however, this depends much on the source material. And I may still have to find the optimum low pass output filtering from the DAC.
Intuitively I've for a long time had this notion that an as-good-as-possible clock would be key to quality/superb digital reproduction. Using the NZ2520SD has confirmed this, although in ways I did not foresee ... I reckon the mind may have a tendency to linearly "project" what it knows already (the new will be "more, better, larger, more beautiful, etc." yet of similar quality) but this to me is a new quality of experience which I'd like to savour in the time to come.
I reckon other clocks may be as good, just different, or even better than the NZ2520SDs in some respects. The ULNOCXOs that have been discussed briefly here on diyaudio could be better, and possibly also this one that I've mentioned previously in this thread:
(http://www.diyaudio.com/forums/digital-source/194965-clock-selection-experts-7.html#post3478464),
however, personally the teaching in this is that (superb) clocks appear to be key to digital music playback - be it the NZ2520SD or another fine clock.
Cheers,
Jesper
I'd like to add a few comments on my experiences using the NZ2520SD in my new (& approaching completion) DAC. The setup is as follows:
Amanero Combo384 -> Acko Reclocker -> own make new NZ2520SD clock board that feeds the DAC and the Amanero (45 & 49 MHz; more comprehensive PSU filtering) -> PCM1794 NOS (DDDAC inspired but different PCB & semi-active output stage; no capacitor). Dual mono altogether.
Impressions on this clock in the new DAC have been "correlated" with a previous PCM1794 version where I briefly had a chance to listen to that DAC with an NZ2520SD compared with the clock coming directly from the Amanero.
My impressions on using the NZ2520SDs: There's really no comparison to other DACs I've listened to in terms of tonal nuance, ambience reproduction, instrument stability, layering of instruments and the ability to discern even small nuances tonally, space-wise, or what could be called a general ability to nuance or "detail". The new DAC with the NZ2520SDs is also very rhythmically stringent: No lagging, sluggishness - just speedy, on time (this, however, I reckon may also be a DAC quality more so than having to do with the NZ2520SD).
To my ears the differences are not subtle ... I tend to compare the NZ2520SD differences with how I've experienced battery powering a full audio system may change the sound - yet battery power cannot "fill in the blanks in the music that are not there" due to a less stable clock ...
The reproduction with the NZ2520SD can also be "sweeter", however, this depends much on the source material. And I may still have to find the optimum low pass output filtering from the DAC.
Intuitively I've for a long time had this notion that an as-good-as-possible clock would be key to quality/superb digital reproduction. Using the NZ2520SD has confirmed this, although in ways I did not foresee ... I reckon the mind may have a tendency to linearly "project" what it knows already (the new will be "more, better, larger, more beautiful, etc." yet of similar quality) but this to me is a new quality of experience which I'd like to savour in the time to come.
I reckon other clocks may be as good, just different, or even better than the NZ2520SDs in some respects. The ULNOCXOs that have been discussed briefly here on diyaudio could be better, and possibly also this one that I've mentioned previously in this thread:
(http://www.diyaudio.com/forums/digital-source/194965-clock-selection-experts-7.html#post3478464),
however, personally the teaching in this is that (superb) clocks appear to be key to digital music playback - be it the NZ2520SD or another fine clock.
Cheers,
Jesper
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.