The BOM for semi-finished board is available at post #3008 on the diyaudio.com thread:
Tempered Master Clock - Building a low phase noise/jitter crystal oscillator
Tempered Master Clock - Building a low phase noise/jitter crystal oscillator
FWIW, when I experimented with the 2020 Driscol, I did not find it to be terribly sensitive to the power supply. I think any well filtered linear PS would do nicely, but defer to Andrea and others wrt the new design.
Thanks Wlowes . I was interested by by Andrea's comment about capacitor on the feed to the clock...deeming them analogue load. And the fact of the sine wave output fair comment. But in the salas thread for whom I have great respect the Refkektor D 'digital' psu is maybe suggested better for clocks due to the low Zo at low Hz such as where the close in phase noise of these clocks is considered.
And then yes I hear 'try your different psu 'and listen a then decide for yourself. And that is the best answer to anyone isn't it really. ...choose for yourself!
And then yes I hear 'try your different psu 'and listen a then decide for yourself. And that is the best answer to anyone isn't it really. ...choose for yourself!
re Power supply
FWIW, the supply I used for 2020 Driscol is a reasonable quality given my limited skills. The clock has separate LT3042 regulators for the crystal and the squarer. The transformer is a medical grade Plitron with separate windings for each supply. I used MUR860 diodes. The clock has a 6th order CRC filter using Panasonic FC caps and an inductor to filter out RF. There are even some Blackgates sprinkled in for good measure. But, it was not a dramatic improvement in sound over a very simple linear supply I used to confirm that it worked. So I concluded good is fine but great not required.
FWIW, the supply I used for 2020 Driscol is a reasonable quality given my limited skills. The clock has separate LT3042 regulators for the crystal and the squarer. The transformer is a medical grade Plitron with separate windings for each supply. I used MUR860 diodes. The clock has a 6th order CRC filter using Panasonic FC caps and an inductor to filter out RF. There are even some Blackgates sprinkled in for good measure. But, it was not a dramatic improvement in sound over a very simple linear supply I used to confirm that it worked. So I concluded good is fine but great not required.
The TWTMC-PPG oscillator is a little more sensitive to the power supply so I suggest a low noise regulator.
Although I have not yet tried the Salas' regulators I believe they are a good choice.
Even LT3042 or LT3045 based regulators.
The DRIXO is the less sensitive oscillator to the power supply, the EXO is in the middle.
BTW, batteries supply would be the better solution for all the oscillators because they will be decoupled from the main AC.
Although I have not yet tried the Salas' regulators I believe they are a good choice.
Even LT3042 or LT3045 based regulators.
The DRIXO is the less sensitive oscillator to the power supply, the EXO is in the middle.
BTW, batteries supply would be the better solution for all the oscillators because they will be decoupled from the main AC.
A newbie question : are the micro voltage made by piezzo electric effect with ceramic class 2 seen in many of those power supply boards are a problem for the TWTMC clock boards ?
I always try to change them when I see one in an analog stage but maybe here it doesn't arm ? (well I know there are usefull when very low inductance is needed due to the small form factor)
@ Andrea, I should have 3 A123 cells lying on a shelf, is the little drift in voltage a bad factor for the clocks circuitry : for instance I noticed they fastly go from 3.3 V to 3.15 or 3.10 V ? Still enough if in serie and regs after ?
I always try to change them when I see one in an analog stage but maybe here it doesn't arm ? (well I know there are usefull when very low inductance is needed due to the small form factor)
@ Andrea, I should have 3 A123 cells lying on a shelf, is the little drift in voltage a bad factor for the clocks circuitry : for instance I noticed they fastly go from 3.3 V to 3.15 or 3.10 V ? Still enough if in serie and regs after ?
Last edited:
Sorry, I have not understood what are you pointing out.
3 cells are not enough, the min voltage for the DRIXo is 12VDC, suggested 15-16.5 VDC.
so what about if I bought more in relation to the question ? 3.3 x 5 = 16.5 and 3.1 x5 = 15.5
so a reg is doing the job whatever the voltage drift with time ? I can rephrase : does a very precise voltage matters or it's ok between 15 to 16.5V
Ceramic cap question is about the numerous ps board one can buy everywhere and is not related to the cells question but more about if micro voltage induced by them as a limiting factor or not if one has to chose such a ps board on internet. Hope my questions are clearer whatever they make sense or not.
Colomba classico, sort of 🙂
so a reg is doing the job whatever the voltage drift with time ? I can rephrase : does a very precise voltage matters or it's ok between 15 to 16.5V
Ceramic cap question is about the numerous ps board one can buy everywhere and is not related to the cells question but more about if micro voltage induced by them as a limiting factor or not if one has to chose such a ps board on internet. Hope my questions are clearer whatever they make sense or not.
Colomba classico, sort of 🙂
Piezzo electric effect in ceramic capacitors class II may occur if the circuit is subject to vibration due to poor attachment, insufficiently dumped chassis and the board itself.
So what can bother the crystal if it is on a vibrating substrate could also bother class II ceramic capacitor.Andrea thought of that too.The crystal is on a dumped pad that prevents vibration and the clock oscillator is housed in its own chassis.
I've seen the inside of a lot of high-end cd players and you'll find such capacitors in all digital circuits.So that's not the real problem.
Regards !
Also diyiggy, keep in mind that all of the performance measurements presented by Andrea are with the BOM specified. FWIW, given the SOTA performance demonstrated, I am not going to question the parts choices.
Yes, you can use 5 x LiFePo4 batteries in series to get around 16V5, that's exactly what we have done with our battery system to supply the oscillators.
Moreover all the phase noise measurements have been done with batteries.
The slow dicharge of the batteries does not matter, the oscillator it's not sensitive to this phenomena.
Finally, there are no regulators on the oscillator boards.
The parts have been selected to get the best phase noise performance, indeed for example we have used ceramic core (Coilcraft) and single layer (Epcos) inductors to get the higher self resonant frequency as possible, low leakage electrolytic capacitor to improve the noise rejection and ceramic (NP0) capacitors where thermal stability was required.
The oscillators and the frequency doublers are RF device and NP0 capacitors are ideal for such application (togheter with silver mica which are larger and difficult to find in non standard values).
Again, I recommend to use exactly the parts indicated in the BOMs.
Hi Wlowes, there is no question with that and I had the power supply boards made for the design. I was asking myself because of the input before mine talking about PS LT3042 boards bought on internet. I experienced they can bring a bad effect, I've been there on IanCanada design with an external LT3042 design and I was worried if not disappointed because the good said reg was giving bad results. It had a lot of X7r at its output and since I experienced with now well known acrylic cap and despite the less good inductance, sounding experience was better ! And it seems many have experienced advices about else caps there I inputed after test for better results, btw you can see anywhere now PPS and acrylic caps as in the FifoPi, etc ! I gave my input about that and it was as old as the classic clock of IanCanada so before the MC and firther Fifo. Details can, could, be important, sometimes ! People should stop paranoia, it's not about theme, their ego diplomas whatelse ?
I think Tom didn't understand my purpose as well. Ceramic class II are bouncing by itself due to the current going through it and I have no doubt Andrea and the smart first advices guys in the main thread are agree about that behavior before I had my first bicycle. 😀. So they know that it's introduce micro voltage drift. I was asking if one want to buy external PS with ceramic class II everywhere as well all see today due to soic regs and their needs and small boards.
Please read the questions as they were without putting in it more than what they ask ! HOW about external PS with such regs and such caps and also with LiPoEtc cells and exact voltage precision as we are talking about sota clocks !
I'm not asking if anyone seen a bush in fire at a top of a mountain or so and questionning about reasonable truth in it ! Nore I am saying anything else that my question is asking ! Please...
It's like some that makes beleive I'm acusating someone while such a people were blowing into my bulocks from many time without prounouncing the words because the form is more important that the deep meaning behind and the attitude! Common !
Though I can recognize my english is basic and sometimes I'm not understood in my questions as I wisch but that's off topic and we don't mind ! So go back to the questions about ones whom would like to powersupply the sota design with SAID external said SOTA boards not from the official BOM just because it is written lt3042 OR WHATEVER... Evil in the details Wolwes, and it's not to you Black Gate enthusiast I can learn anything here ! 🙂
Bad to waste a good design with choosing maybe not a good link working with... that the sense behind my questions !
Last edited:
Thanks Andrea for the answers.
Yep I have the boards for PS and am confident with the BOM, after all can see some Pan FC in it 🙂
Questions was asked also for others as explained in the post above !
This afternoon I was drinking a Coteaux du Layon (a white sugared wine for ones that don't know) with an italian Paques cake while contemplating the board... there are nice moments in life, thanks for that !
Yep I have the boards for PS and am confident with the BOM, after all can see some Pan FC in it 🙂
Questions was asked also for others as explained in the post above !
This afternoon I was drinking a Coteaux du Layon (a white sugared wine for ones that don't know) with an italian Paques cake while contemplating the board... there are nice moments in life, thanks for that !
diyiggy
I mean no disrespect and just offered my thoughts. Recall that you and I collaborated on our positive experiences replacing ceramic caps on the FIFOpi clock adapter pcb so this was really just a long running conversation. Probably contributes little so just ignore me. If you discover a nugget of information I'll be the first to appreciate it.
I mean no disrespect and just offered my thoughts. Recall that you and I collaborated on our positive experiences replacing ceramic caps on the FIFOpi clock adapter pcb so this was really just a long running conversation. Probably contributes little so just ignore me. If you discover a nugget of information I'll be the first to appreciate it.
@diyiggy: Hi ... Just a brief comment wrt LiFePO4 batteries: If you intend to use them in series connection, and also charge them in series, may I suggest that you implement some kind of over-voltage protection per cell, as they tend to become imbalanced over time. Such an imbalance may cause some batteries to experience over voltage which may diminish their life span significantly.
Cheers,
Jesper
Cheers,
Jesper
Sonic impressions
Hi,
My new Driscoll 5,6mhz clock unit is running since last Sunday and these are my first impressions.
The setup:
The clock is powered with a battery board from Ian with 4 batteries in series and ultracapacitors.
It is clocking a USBridge signature with fifopi feeding my TDA1451A dac, build with salas regulators and a tube IV stage into directly heated triode amps (using 20A-interstage-205D-interstage-AD1 in balance mode-output transformer. EML and Psvane WE replica tubes, Tamura transformers) into open baffle speakers with partly self build field coil wideband units (no filter!) and Phy tweeter.
This is an extremely revealing set that lets you hear every small change, if I only reheat a soldering the difference is already audible (this can be quite a pain in the....)
All music files are 16-24/44,1khz as this is the limit with the 5,6mhz clock.
The comparison is with the already excellent previous driscoll clock from Andrea with SC-cut 45mhz oscillator.
Impressions:
Most apparent is that the musical details in the sound seem to be more compact, fixed in place and better timed. This makes it easier to hear all these details, there is more insight.
You can more easily hear the atmosphere of the recording room. But also tape hiss on old recordings is more detailed......
The S and T sounds of voices is less pronounced, this is already very good with the old driscoll but is now on a higher level.
The musical flow is very nice.
There is not any trace of a digital sound, everything is sounding natural and honest. But sometimes too honest, also all faults in the recordings are clearly audible.
Did I mention the dynamics? The dynamic is improved, delivering more focus and more pulse.
It is so easy to listen and get into the flow, no need for high sound pressures, the volume is now on a lower setting and I do not miss anything in the presentation.
Is this SOTA? Yes I think it is, I have never heard any system sounding as revealing as this one.
If you reach this level and experience how good the old red book material is, you will forget about the high sampling rate marketing hype that need high clock frequencies with accompanying higher phase noise. Why bother if red book sounds so awesome!
Conclusion: very happy with the result!
Note that this analysis is done with my system, your experience may be different.
Next up will be to build the rest of the set: 6mhz, 22mhz and 24mhz driscoll and 4x doubler
Regards,
Hi,
My new Driscoll 5,6mhz clock unit is running since last Sunday and these are my first impressions.
The setup:
The clock is powered with a battery board from Ian with 4 batteries in series and ultracapacitors.
It is clocking a USBridge signature with fifopi feeding my TDA1451A dac, build with salas regulators and a tube IV stage into directly heated triode amps (using 20A-interstage-205D-interstage-AD1 in balance mode-output transformer. EML and Psvane WE replica tubes, Tamura transformers) into open baffle speakers with partly self build field coil wideband units (no filter!) and Phy tweeter.
This is an extremely revealing set that lets you hear every small change, if I only reheat a soldering the difference is already audible (this can be quite a pain in the....)
All music files are 16-24/44,1khz as this is the limit with the 5,6mhz clock.
The comparison is with the already excellent previous driscoll clock from Andrea with SC-cut 45mhz oscillator.
Impressions:
Most apparent is that the musical details in the sound seem to be more compact, fixed in place and better timed. This makes it easier to hear all these details, there is more insight.
You can more easily hear the atmosphere of the recording room. But also tape hiss on old recordings is more detailed......
The S and T sounds of voices is less pronounced, this is already very good with the old driscoll but is now on a higher level.
The musical flow is very nice.
There is not any trace of a digital sound, everything is sounding natural and honest. But sometimes too honest, also all faults in the recordings are clearly audible.
Did I mention the dynamics? The dynamic is improved, delivering more focus and more pulse.
It is so easy to listen and get into the flow, no need for high sound pressures, the volume is now on a lower setting and I do not miss anything in the presentation.
Is this SOTA? Yes I think it is, I have never heard any system sounding as revealing as this one.
If you reach this level and experience how good the old red book material is, you will forget about the high sampling rate marketing hype that need high clock frequencies with accompanying higher phase noise. Why bother if red book sounds so awesome!
Conclusion: very happy with the result!
Note that this analysis is done with my system, your experience may be different.
Next up will be to build the rest of the set: 6mhz, 22mhz and 24mhz driscoll and 4x doubler
Regards,