Andrea
I am not asking you to measure it. There are a series of test specs provided by Laptech that you share with the GB. But I cannot find any of these specs used to describe the Crystek or NDK SDA devices that many of us are using. So I thought that perhaps you or one of the other experts would be able to compare and contrast the performance spec of the Well Tempered Master Clock vs these commonly used products as it pertains to use in an audio DAC.
I will draw my own subjective comparison of what it does to the sound with my ears in my room, but it would be nice to have some meaningful data to be in a position to understand the difference in sound I hear (or do not hear).
If that is not possible, don't worry about it.
I am not asking you to measure it. There are a series of test specs provided by Laptech that you share with the GB. But I cannot find any of these specs used to describe the Crystek or NDK SDA devices that many of us are using. So I thought that perhaps you or one of the other experts would be able to compare and contrast the performance spec of the Well Tempered Master Clock vs these commonly used products as it pertains to use in an audio DAC.
I will draw my own subjective comparison of what it does to the sound with my ears in my room, but it would be nice to have some meaningful data to be in a position to understand the difference in sound I hear (or do not hear).
If that is not possible, don't worry about it.
Thanks, Andrea. I don't think these variation is big enough to induce hearing difference. I might be wrong though.
Keep in mind that Laptech sells crystals and gives you the specs for those vs Crystek and NDK sell you oscillators that contain the crystal and circuit inside so they won't bother to tell you the spec of the crystal itself. They don't even give you the phase noise plot down to 1Hz.
Also, it seems that there is quite a bit of variation in the quality of at least the NDK oscillators to an extent that vendors buy a big batch then measure each one to get the ones with better specs and sell the rest for lower price.
Thanks Wealas
I celebrate the work that Andrea has done to build the Well Tempered Master Clock. I have accumulated the parts and over the next couple of weeks expect to hear for myself the results of this ambitious undertaking.
I am concluding however that we are all taking a leap of faith without really knowing how the specs of the finished master clock compare with commercial offerings designed to serve the same function. This does not particularly concern me as I will judge the expenditure of time and money based on the impact of the sound I hear. I now understand that measuring the device is not a trivial task. I see why there is so much confusion in the audio industry about the role of the clock. I see reviewers of high end audio waxing poetically about the spec of the clock in the $30k Esoteric DAC and in their $20k master clock. Then I learn the specs they quote have nothing to do with sound quality. Big surprise
I celebrate the work that Andrea has done to build the Well Tempered Master Clock. I have accumulated the parts and over the next couple of weeks expect to hear for myself the results of this ambitious undertaking.
I am concluding however that we are all taking a leap of faith without really knowing how the specs of the finished master clock compare with commercial offerings designed to serve the same function. This does not particularly concern me as I will judge the expenditure of time and money based on the impact of the sound I hear. I now understand that measuring the device is not a trivial task. I see why there is so much confusion in the audio industry about the role of the clock. I see reviewers of high end audio waxing poetically about the spec of the clock in the $30k Esoteric DAC and in their $20k master clock. Then I learn the specs they quote have nothing to do with sound quality. Big surprise
I would have thought many people here have access to phase noise test kits.
I have HP3048A and E5500 - if you want to send me a few samples I will measure them for you out of curiosity and send them back if needs be.
Ideally - two devices of the same type so that they can be tested against each other (with 3dB adjustment.)
Cheers
Leo
The new comin Discroll from Andrea should hzve even better specs.
I noticed improvment by decoupling a cchd 957 with acrylic 0805 smd caps, could it be better than a good X7R ? Or had I luck of a good mix with my dac ? That said wlowes member also tried and found an improvment with NDK as well...
If the new discroll board have some 0805 smd traces needing 0,1 uF or more, I will compqre two boards, one with ceramic and the other with ceramic caps...in the pico range ceramic is the only choice though.
I noticed improvment by decoupling a cchd 957 with acrylic 0805 smd caps, could it be better than a good X7R ? Or had I luck of a good mix with my dac ? That said wlowes member also tried and found an improvment with NDK as well...
If the new discroll board have some 0805 smd traces needing 0,1 uF or more, I will compqre two boards, one with ceramic and the other with ceramic caps...in the pico range ceramic is the only choice though.
Hi Leo,
thanks for your support.
Unfortunately I have no finished oscillators to send you for testing.
Does anyone have a pair of oscillators at the same frequency to send to Leo for testing?
Andrea
High praise for WTMC impact on sound
Well, we may not have the jitter spec nailed down, but I confirm that the impact on sound is fantastic. I have had the 45mHz clock running now for a few days. Certainly takes the DAC to the next level. Thanks again Andrea for sharing this innovation. I'll comment more fully in a few weeks when all finished and burned in.
Well, we may not have the jitter spec nailed down, but I confirm that the impact on sound is fantastic. I have had the 45mHz clock running now for a few days. Certainly takes the DAC to the next level. Thanks again Andrea for sharing this innovation. I'll comment more fully in a few weeks when all finished and burned in.
Attachments
Thanks for sharing your impression, I'm very glad you got a sound improvement.
I look forward for your detailed comments.
Hi,
AKA : PPS caps (smd package): look for Cornell Dublier, Panasonic, Rubycon...
They don't sound better but give a different sounding result than ceramics which can be usefull sometimes according to the equilibrium of your devices listened as a whole (so subjective as checked with ears)
Excellent behavior at high frequencies in class with ceramics class A. Default is package is bulkier (so not better inductance), precision is max 2% (Panasonic). I founded it usefull to coock the CCHD 956 oscillator that have small ceramics values on board : putting 1 uF between the best powersupplies (noise & PSRR) has given a better aesthetical result (tested on Iancanada clock boards in lieu of ceramics I didn't like)
I asked wlowes to check it on NDK crystals, he had also good result (read: better at ears) following the way I did (step by step with listening checks to find the right amount of capacitance); he swapped the ceramics for PPS while keeping BG NX caps.
So PPS are quite good for decoupling but where capacitance must be very precise (no pF range & not better than 2% on few values) or inductance very small (no package below 0805 size). Worth it in my setup between 0.1 uF & 1 uF. I found (sorry to let cry some people, I don't want to open a worms can here) Cornelle Dublier (precision 20% only iirc) to give better subjective sounding result that the Panasonic equivalent. Didn't trie the Rubycons yet.
IMHO : to use when you can swap any ceramics class II (and I if no precision & low inductance needed) 0805 and highest form factor sizes. Especially in HF circuitries.
hope that helps. Better here to try than theorizing... I discover it thanks to the long thread of ECDESIGNS and saw also it is used by a guy famous to let sound TDA1541A at his best, ( Pedja Rogic from Audial, he offers diy pcbs as well), and give them a chance for the better in my setup.
Last edited:
Thanks for your opinion on these. If they're being used in the PS then no need to get any better than 2%. I can understand using them over ceramics in the analog section but in the digital section and high freq apps I'm not so sure. Do you have a link to the section of the ecdesigns thread where they are discussed.
I was also surprised as it's far above what we can hear. Maybe the difference is placebo, but when you check with family or friends without saying anything they also notice something changed. I had also excellent result, far better than ceramic on an ESS dac chip for low noise ldo feeding the PS as also decoupling for charge pump, reference.
Or maybe our circuitries are ringing an such caps fix some things, I really don't know. But if your hifi gives a thin sound, with fatiguing highs, not so articuled bass, such a caps gave me the opposit in the same circuitry despite it was in the digital domain.
Btw, off topic, but iancanada putted a big Rubycon acrylic around 15 uF iirc in its late dev. : FifoPi. I had this improvment on the older clock board.
I think it's really how the whole hifi sounds, you may or you may not like the change, but you should notice a sounding difference. Now how it is on a perfect development made with 100% ceramics around the DAC ? I don't know and I will be not surprise the swap with acrylic give not a better result! Maybe it is just giving more damping than ceramic ? I can't tell.
I see it as a tool to setup the sound as I could add some pepper. Anyway it worked with Crysteq CCHD-957 despite a fifo and a clock board with sota and standalone power supplies feeding a dac via uf-l wires via I2S protocol. I know it's not a rational talk but two 0805 smd caps of the same capacitance value: X7R vs PPS gave me different sounds and it was not a subtle change.
Don't know how it performs in an analog circuitry, never tested in serie as well !
The best is to give it a try to see if it's ok or not on a small board like the Andrea's Discroll.accoring to your whole setup...easy to try.
Or maybe our circuitries are ringing an such caps fix some things, I really don't know. But if your hifi gives a thin sound, with fatiguing highs, not so articuled bass, such a caps gave me the opposit in the same circuitry despite it was in the digital domain.
Btw, off topic, but iancanada putted a big Rubycon acrylic around 15 uF iirc in its late dev. : FifoPi. I had this improvment on the older clock board.
I think it's really how the whole hifi sounds, you may or you may not like the change, but you should notice a sounding difference. Now how it is on a perfect development made with 100% ceramics around the DAC ? I don't know and I will be not surprise the swap with acrylic give not a better result! Maybe it is just giving more damping than ceramic ? I can't tell.
I see it as a tool to setup the sound as I could add some pepper. Anyway it worked with Crysteq CCHD-957 despite a fifo and a clock board with sota and standalone power supplies feeding a dac via uf-l wires via I2S protocol. I know it's not a rational talk but two 0805 smd caps of the same capacitance value: X7R vs PPS gave me different sounds and it was not a subtle change.
Don't know how it performs in an analog circuitry, never tested in serie as well !
The best is to give it a try to see if it's ok or not on a small board like the Andrea's Discroll.accoring to your whole setup...easy to try.
Different applications will bring different results. Just like you'd never use ceramics in an analog circuit, well at least I wouldn't. If you are using these PET caps for the multiple decoupling caps in your TDA1541 then okay I would agree they would sound better than a ceramic. Using them in a high speed digital circuit PS bypass then I would disagree. Using them around the crystal circuit it would needed to be ab'd, using them on the squarer circuit I would suggest the ceramics might be a better choice. How you perceive the resulting sound depends on the rest of your system as you don't listen to it in parts. After all it's only as good as the least perfect component(ie: amps, speakers-drivers etc).
And everyone has what they like to call the perfect sound and unless you've listened to live unamplified concerts then ones perception of what good sound is could be skewed to start with. A simply test is go to a music store and play with a cymbal using different drums sticks and get an idea of what they should like then compare that to your system. Play a variety acoustic guitars to hear their sounds etc.
And everyone has what they like to call the perfect sound and unless you've listened to live unamplified concerts then ones perception of what good sound is could be skewed to start with. A simply test is go to a music store and play with a cymbal using different drums sticks and get an idea of what they should like then compare that to your system. Play a variety acoustic guitars to hear their sounds etc.
I know and I can't agree more.
But is it important we agree each other ? The best is to check the knowledge with experiments, agreing or not without a try is just for living room or kitchen talks. Perhaps I had just some luck as wlowes fellow had with it and a // BG NX, maybe we are all biased and never listened music... who knows ! Maybe we are in primary school yet
I don't know what you call high speed, 50 Mhz to 100 Mhz for ESS chips is for me more middle speed ! Well, give it a try with such dac.
But is it important we agree each other ? The best is to check the knowledge with experiments, agreing or not without a try is just for living room or kitchen talks
. Perhaps I had just some luck as wlowes fellow had with it and a // BG NX, maybe we are all biased and never listened music... who knows ! Maybe we are in primary school yet I don't know what you call high speed, 50 Mhz to 100 Mhz for ESS chips is for me more middle speed
! Well, give it a try with such dac.- Status
- Not open for further replies.