Since my trip to Peters I have been working to understand why the dexa usb clocks did not perform as expected. The first step in understanding what happened is in this post.

http://www.phasure.com/index.php?topic=2888.msg30710#msg30710

In summary this post flags the sound quality performance difference that tuning the PC and DAC usb clocks makes to sound. Being a compulsive tweaker, when Peter and I listened to the clocks I had the damed clocks significantly out of sync (of the order of 1000s of hz against a USB chipset spec typically of 24mhz +- 100ppm). It was only after returning from Peters that I worked out how to tune clocks accurately to each other.

I the post above I had to stopped short of saying that the poor performance we encountered was completely explained. This was because although synchronising the USB clocks makes a massive difference to sound I still felt there was something else about the set up that had impacted sound. So more thinking and now what i beleive to be the final issue solved making the dexas perform as i had expected.

What had happened is that whilst tweaking, I had set the speed of the NOS interface USB clock fast by 1-2 kHz and lost track that I had done this ! I was stumped as to how I could get the NOS dexa close to dead on 24mhz again (my scope frequency counter is only accurate to 3 decimal places), then I remembered the DIY 24mhz clock I'd built to test the USB clocks  . Using the diy clocks as a reference the dexa at the NOS end was retuned again to very close to 24mhz. Then PC end dexa usb was re-tuned to match the DAC end dexa with both running much closer to 24mhz.

So now the system behaves exactly as expected (totally different performance) without question these two tuning factors are the root cause of the poor performance that Peter and I experienced.

I'm really glad to have found the cause of the problem, diagnosing what happened has helped to pin down a better understanding clock related variables and how usb clocks need to be set up. It would be fun now to retry the upgraded clock config in Peters system but not that practical unfortunately 

Anyway absolutely top, top draw performance again from the modified USB link.

Regards,
Nick

Hi Peter,

I guess I wanted to highlight that in my and Pauls case (with Dexas) even with the clocks "tuned" there is cycle of change happening in the relative Phase of the clocks. (0 to ~5 beats per second is the best I can manage at the moment, its about bang on the 0.2ppm “precision” spec Dexa quote). The timing shift from the clocks drifting is similar in order of magnitude to the 10ns static timing change from the cable. I just wanted to make sure that there was not an assumption that the clocks did not drift with respect to one another and that the only timing change was the cable length.


My thinking is similar here I think regards the current and noise balance, its interesting that the dac clock running much faster / slower than the PC changes sound quality in different ways. Could this be linked ? I still have the suspicion that the "lock" circuit of the receiving end (via on chip PPL ?) may find it easier to maintain its lock on the incoming data when the clocks are close or for that matter running faster or slower with respect to one another.


The display of relative speed is continuous without “hops”, it can resolve <5 degrees of relative phase shift without having to try to estimate two separate clock wave forms by eye. Its not good at measuring large "beat" rates between the clocks but when the clocks are closely matched (10 beats and less per second) its easy to see the speed of relative phase shift. The neat thing is that the trace can be watched easily whist playing music from the listening position, so clock +- relative movement and speed can be listened to, really nice.


This concern was just when an attempt is made to measure an Xtal output directly. I'm guessing that the 10pf or so of probe capacitance will change the Xtal resonance point and change the measured frequency. This is why I don’t think its possible to correctly measure / observe the clocks on unmodified PCIe and NOS1 boards. In the case aftermarket clockes with buffered outputs I don’t think the scope probe capacitance should matter.

Regards,

Peter, Paul hi,


I don't think the effect that Paul is describing can be linked to a "static" shift in USB clock or USB data phase of 10ns introduced bt the shorter USB cable.

I agree that the cable length will introduce the static 10ns change in USB signal timing, but what you may not be aware of is that there is already constant dynamic relative phase drift of the two USB clocks and that the levels of drift are of the order of the 10ns static change transmision time.

The dexa clocks when tuned go through a cycle of drift relative to one another (description in the the link below). Pauls clocks behaved identically to this when I set them up at my place.

http://www.phasure.com/index.php?topic=2888.msg30714#msg30714

A good estimate from tuning with my 'scope is that the rate of change of the phase angle between the two 24Mhz USB clocks varies between 0 degrees per second (clocks running at the same speed) to 1800 degrees per second (clocks running out of sync by 5 cycles per second). Each cycle of the 24mhz clocks is 40 nano seconds in length which means that the relative speed (and phase relationship) of tuned dexa clocks is of one clocks is moving constantly from synchronisation positions of 0ns to 200ns then back to 0ns.

The D+ D- USB carrier dynamic phase change as a result of the clock drift "could" be much greater as a result of the clock phase changes. The carrier frequency is 480Mhz so a carrier cycle time is a 1/20th of the USB clock cycle at 2 nano seconds (I know we did not agree on this when we were discussing this data speed, but take a look at the “Signalling” section of the link below).

http://en.wikipedia.org/wiki/USB#USB_2.0

“A high-speed (USB 2.0) rate of 480 Mbit/s was introduced in 2001. All hi-speed devices are capable of falling back to full-bandwidth operation if necessary; i.e., they are backward compatible with USB 1.1. Connectors are identical for USB 2.0 and USB 1.x.”

The point is that there is already constant “change” in relative phase of the clocks that is of the same order as the 10ns timing change introduces by Pauls shorter cable, so I don't think that introduction of a “static” 10ns timing change from the change in cable length is likely to be the cause the Pauls effect.

Now I'm just throwing out ideas here, but could be that the shorter length USB cable somehow happens to have better signal reflection characteristics or lower signal smearing for the USB signal ? or could the signal earth carried by the USB cable from the PC be "better" coupled to the DAC by the shorter run of cable ?

Regards,

Nick.

Georg,

I think you could be bang on with what im seeing here.

The dexas have a simple resitor heater and thermostat control of the temperature of the 24mhz xtal can on the clock pcb. When the clocks are turned from on cold there is a very large difference in relative speed of the clocks. This reduces quickly (as the temps stabilise I think) at first quickly then further stabilises happens over about 10m. In the end the two clocks do not stay completely stable with respect to each other. I've got them trimmed now with a slow progression of 0 beats per second to around 10 beats per second (possibly less). The cycle from 0 beats to 10 beats and back takes about 30 seconds. I cannot prove it right now but I have a feeling this might be due to the heating cycle of the thermostat on both clocks slight over heating then under heating the crystals.

Thinking about it the beat frequency cycle could be due to other effects as well. But hysteresis in the heating / temp sensing circuit seems like it might be the first place to investigate. It's fun this stuff, whats the betting that Paul's clocks behave completely differently when we look at the haha :-).

Best,

Nick.

Coen hi,

I agree this needs to be tried. I have had all the parts for a while now to give a single clock driving both ends of the USB link a go, just need to find the time  .

Regards,

Nick.

Alain, All,

After returning from my visit to Peter's I was left with a big question - Why did the USB clocks not perform as expected in Peters system ?

Peter provided an account and his view above in this post of what may have been happening regards USB clock noise etc. I know that the dexa clocks (read improved clocks on standard xtals in USB interfaces, forget Dexa as a make here) have made a significant positive impact in 5 of the 6 systems that they have fitted. Before posting my comments and thoughts on how the USB clocks performed at Peter’s and why, I wanted to spend time investigating what might have happened to impact their performance. I have been doing this for a couple of weeks now and am reasonably sure that I understand now what happened. I will post more on this shortly, but for now…

To the point that you make above Alain, the relative speed of the two clocks at the ends of the USB link is very important to sound quality. As part of the work to diagnose the problems at Peter’s I now have a way to show the relative speed of the two USB clocks in my system. This is done is such a way that the relative clock speed can be looked at in real time across the room from my listening position whilst listening to music. The display will resolve fractions of one cycle of phase shift between the clocks (eg a few degrees of one 360 degree “beat” between the clocks). Viewing the relative clock speed in real time whilst listening its clear that speeds do significantly impact sound quality. In fact it changes sound differently if the PC PCIe USB clock is running faster or slower then the NOS1 USB interface clock. Same speed, or as close as possible to this, is best subjectively for SQ.

The USB clocks in my system are tuned now to a frequency which is better than 1 part per million of each other (this 1ppm is the max and for some time they run at the same speed exactly). When we tried the clocks at Peters they were matched very badly (worse than 100 parts per million). Sound now is entirely different to what Peter and I were listening to.

If these results are valid and relative USB clock speed does impact sound quality, then without the capacity to tune the USB clocks it may very well just be luck that determines how good a particular PCIe USB3 card sounds with a NOS1. The luck being that all the many variables that impact the Xtal speed and stability of on the USB card and NOS1 USB interface just happen to “line up” such that the two crystals run at close speeds to one another and sound good. That is a worrying thought as even different instances of the same PCIe USB card model may sound different…..

Paul is bringing his Silverstone USB card, and DAC (fitted with USB Dexa clocks) to mine this weekend to look at this and we hope to get more supporting evidence. With respect to the SIlversone USB card, unfortunately the act of measuring the oscillation speed of a raw crystal’s output will change the speed (due to capacitance of the probe that is being used to do the measuring) so tuning the card to the Dexa in my NOS1 is unlikely to be possible but we will see what we can do. I will update when we have more information.

Regards,

Nick.

Thanks Brian,

I just wanted to make sure that everyone is aware of the comparison. I would hate for anyone to invenst in the 22.5 or 24.5mhz dexas for the NOS after what we found.

Actually Paul has also helped reaching thisc onclusion as well. A couple of weeks before spending the time with Peter, Paul came over to mine and we had a chance to compare our DACs, mine with a dexa 22.5mhz clock and Pauls with the standard clock module. This also cast doubt on the dexas relative performance.

At Peters listened to under the frankly microscope performance of Peters system (WOW its good) it was clear that the standard module was just more musical.

I'm happy that this is the case, over the years I've spent too much on expensive aftermarket clocks for equipment so it's nice when a properly implemented module at a sensible cost sound better :-).

In the area of 24mhz usb clocks "the jury is out" just now. They defiantly did not help at all in Peters system but they are still a big upgrade here in my system and I think in a number of other folks, so ill be looking carefully again at them, but it just might be that their effectiveness is in some way system dependant. I guess we will see.

Best,

Nick.

Hi,

I just want to update information above regards using a dexa clock to replace the 22.5 MHz audio clock of the NOS1. I have also copied the following text into the first post in this thread so that the comments are also in context of the original post.

I had the privilege of visiting Peter for a couple of days last week. Whilst at Peters we had the chance to listen the NOS standard 22.5mhz clock and the dexa 22.5mhz clock to get a feel for how they both sounded. In addition to this Peter used his FFT measurement equipment to measure the the audio band psu noise of the NOS clock and the dexa clock.

This all made for some interesting results which I need to update here in order to keep information current.

After listening to both dexa and the NOS standard 22.5mhz clocks, the NOS standard clock provides the best sound quality.

The audio band FFT traces of the power supplies to the dexa and NOS1 clock showed the NOS1's clock supply to be around 45db quieter than the dexas psu. This is not a direct indication of the clocks performance but it does show how the lengths to which Peter has gone to give the NOS1 clock the best environment to perform in.

In summary the NOS1 standard clock sounds better than the dexa used as an audio clock. Stick with the standard clock it works exceptionally well !

Cheers,

Nick.

David hi,

I know your question was to Brian but I though I would chip in.

I think many people have found that using a NOS1 makes it possible to get a really great sound from quite modest speakers and amplifiers. It seems to take things back to the system hierarchie of the 80s and 90s when it seemed every hifi shop would tell you to spend most of your budget on the best source you could (turntables back then). The NOS1 really works in this way. I have had really surprising and enjoyable sound driving a 5.1 computor audio system for instance with my NOS1 :-). I think there are a number of stories from NOS1 owners who have been very surprised by how good sound is from "modest" amps and speakers (usually this happens when "main" systems for some reason are not working)

I think provided your amp and speakers are reasonable (and the reviews I looked at seem to say they are) then putting a NOS1 on the front end is likey to give really good results. Long term, once it's in your system you can upgrade your amp and speakers (if you want) and you can be sure that the NOS1 front end will live with what ever you put in.

A couple of points to consider are that you can bypass your amps volume control (if you have a tape loop input on your amp) which can be good for sound. You may also want to take a look at your PC build just to check that it will drive the NOS1 effectively.

My vote would be go for it, I very much doubt you would be disappointed with the results.

Regards,

Nick.
 

Hi,

I’m really quite excited about this post, I was going to add this post to the “hunting for noise” thread as a packaging addendum to the clock posts in that thread but what has happened here draws on thoughts from a number of other recent topics and discussions regarding grounding, EMI susceptibility and clocks, so here is a new post.

If you buy into the need to drive audio circuitry with low jitter / noise oscillators and think for a moment about the challenges of generating a “clean” clock wave form then clocks are debatably one of the most EMI susceptible elements of the replay chain. For about a year now my DAC has used a non-standard audio clock, then two upgraded USB clocks were added to the system. Two of the clocks ended up perched precariously on a plastic sheet resting on top of my NOS1 box. This is not because they have not been mounted with the NOS1 case, they have, but because with the resolution of the NOS1 and the rest of the system you can very easily here negative impacts on sound quality arising from placement of both the clock circuits and their associated power supplies. The PC USB Clock is equally sensitive, both Paul and I have posted about changes in sound as this unshielded clock is moved from location to location. Based on previous experiences using third party clocks and recent experiments reported in the “Hunting for Noise” thread it has always been the intension to individually fully (well almost fully there are a couple of holes in the enclosures) shield all three clock PCBs and their PSUs. Over the last month or two the first stage of this work to shield the clocks has been completed. The PSUs are still to be finished if the back order for the enclosures that have been ordered ever gets delivered !!

From a packaging point of view to get two of the clocks (audio and USB) to fit into my NOS1 case (Im desperate to get the top back on again !), it meant choosing quite compact metal enclosures for the clocks. Its been a challenge to make the clocks fit and in the end it was necessary to change components on the clock PCBs to make them fit into the small enclosures.

Very careful attention has been paid to the grounding of the clock PCB, PSU and in particular the grounding of the enclosures that the clocks are now inside.

The results.


What follows is with the Audio and USB clocks now installed within the NOS1 case and the PC USB clock fitted outside of the PC chassis.

First sound quality change in summary is very significant indeed, I would place the improvement in the region of 1/3 of the step that all of the modifications the my NOS and PCIe USB card have made so far, so very significant indeed. The clocks have indeed proved to be very sensitive to shielding. Just how much of difference it makes has come as a real surprise. It’s no secret that I’m a real believer in the merits of the upgrading USB clocks at both end of the USB Link bit this has moved things to a new level again.

DAC Audio Clock -see Update below


This is where things get interesting. The baseline position was that I personally preferred the Dexa clock that I have been using subjectively to the NOS1’s oscillator module but I was happy to accept that the difference was not that great and was likely to fall into the “personal preference “ type of change for folks listening to the two side by side. Certainly those that have heard these clocks back to back have noticed this difference but stopped short of investing. With the shielding applied however  there is a marked improvement in sound quality I would now have to say that there is a gap between the performance of the non standard clock (now a clear a favourite for me) and the standard clock in terms of subjective sound quality.  It will be very interesting for other folks to re-listen again as see what they make of this.


UPDATE 2/3/14

I have had the privilege of visiting Peter for a couple of days last week. Whilst at Peters we had the chance to listen the NOS standard 22.5mhz clock and the dexa 22.5mhz clock to get a feel for how both sounded. In addition to this Peter used his FFT measurement equipment to measure the the audio band psu noise of the NOS clock and the dexa clock.

This all made for some interesting results which I need to update here in order to keep information current.

After listening to both dexa and the NOS standard 22.5mhz clocks, the NOS standard clock provides the best sound quality.

The audio band FFT traces of the power supplies to the dexa and NOS1 clock showed the NOS1's clock supply to be around 45db quieter than the dexas psu. This is not a direct indication of the clocks performance but it does show how the lengths to which peter has gone to give the NOS1 clock the best environment to perform in.

In summary the NOS1 standard clock sounds better than the dexa used as an audio clock. Stick with the standard clock it works exceptionally well !
   


PC USB Clock


Proper shielding makes a huge difference in performance particularly at the PC end. It’s worth pointing out that prior to putting the USB clock into its new enclosure and setting up its grounding the clock was not living inside of the PC (eg in harms way for EMI as it were). Rather it lived outside of the case in the best location I could find for sound quality. With the enclosure and grounding scheme in place there are more options for location of the clock now without impact on sound. And sound is just superb, it is one of those changes where everything improves from the lowest of lows to the highest of highs and everything between. Bass in particular is very improved, not overstated at all just much much better there is resolution way way down now and low “bass effects” mixed in during production that were not easily audible are playing along tunefully now.

DAC Interface USB clock


Again a significant difference in sound quality although not so large as that gained from shielding and grounding of the PC end USB clock. Similar sound quality gains characteristics but slightly less pronounced defiantly worth having though.

Thoughts.


Firstly as anticipated to sweat more performance out of the clocks it’s very well worth while taking precautions to prevent EMI.

Changes to the clock shield / PE electrical linkage for each individual clock changes sound characteristics. Get this “right” and sound beautifully clicks into place.

The behaviours from these experiments with the shielding strongly point to the PE grounding practice that Mani and Paul have implemented whereby each main component PC, DAC, AMP have their own PE spur running in star ground fashion to a dedicated PE spike. Following their PE star setups was on the plan after listening to Mani and Pauls systems but this has shot right up in priority after playing about with clock shielding. NICE ONE GUYS !

Finally it will be interesting applying enclosures to the three Clock PSUs which are currently still unshielded. I would not anticipate quite the same level of impact on sound quality but who knows.

As a final foot note.


The intension here is to highlight the susceptibility data and audio clocks to EMI and the effect on sound quality this can have. I guess to some extent or other it will be the same if the clock is a DIY module build, a standard XTAL or XCO fitted directly to  PCB or a 3rd party product such as the PPA card or Dexa. It’s just good to understand that whatever the clock is, the conditions it operated in might REALLY matter for sound quality.

As usual no measurements here just subjective assessment of sound quality changes, although differences in sound are so significant that conclusions about the effectiveness are easily formed.

Best Regards,

Nick.

Mani hi,

My hunch would be to go with option 1. The reasoning being that it can use short interconnects. I tried 15m usb cables and my hunch is that a long USB lead would be less disruptive to sound than 5m + interconnects. There would be a lot of dielectric length in the ICs to "shape" the sound and to manage EMI on such a long cable it will need to be shielded which might make keeping capacitance low difficult.

If it were me, I would also look to reduce the length of the USB cable by  moving your PC into your listening room. I think you were running fan-less iirc when we listened to your system so with a smaller PC case it would not be too disruptive to your listening room.

With the bass output of the Orelinos that Peter has described, if the NOS / PC were to move into your listening room I was thinking it might be useful to cut structural vibration from the suspended wooden floor by finding a way to put in a wall mounted shelf to use the mass of the house to isolated bass vibration for your NOS / PC. Similar to the set up on your studio turntable support.

That leaves airborne vibration which I think could still be a factor. But I think it may be that reduction of USB and IC length whilst using wall based platform to cut structural vibration might help address the factors that have the greatest impact on sound.

My set up here is reasonable close to what I'm suggesting above, although I don't have a wall attached platform my Townsend rack is placed on the concrete floor.

Which ever way you go it will be really interesting seeing what the relative influences and impacts are on sound quality.

EMI is certainly a factor but i suspect that it may be in key areas of the playback system. I got round to doing some planned targeted work on EMI reduction over the last couple of months and have had some absolutely superb results from it. I'll put a post up shortly about it (all applicable to you modified DAC).

Kind regards,

Nick.

Hi,

Iv been following the thread but had decided not to say much, I'm not familiar with the silverstone card and although the stuff I tried (below) was not conclusive that's not to say that there were not changes to sound. I found results to be quite system dependant so I did not want to comment too much because as it seems likely that we may experience the effects differently depending on the rest of the system.


IV destroyed a good may usb2 and 3 cards over the years trying similar things  Rather than use a pcie riser I traced the pins on the usb PCIe edge connector back to the bus pins and broken traces to inject power (LPS, batteries etc), disconnect PCIe earth and decouple cards from PC chassis earth, and DAC signal earth. Also tried clock mods and usb card component upgrades. Then there was galvanic isolation with the Anadco.  In the end it was thinking about results of experiments with the Anadco that lead back to experiments with usb clocks again.

Regards,
Nick.

Hey Anthony,

The datasheet of the TI did look interesting but the chip takes a 1.somthing volt clock input so id need to attenuate the dexas output or build a test clock using a module with the correct output.

I guess trying the TI chip on a standard card against an NEC would give a good idea if the TI  shows promise.  Right now its difficult to see how the nec / dexa with todays setting up work applied could get better, but how many times have we thought that before but that's the fun of this hobby.

Your right though, the TI based card needs to be tried.

Cheers,

Nick.

Hey Paul,

That's one less job to do here then!

I would guess that the Crystal on the Silverstone card is a little better than on other standard cards. Remove and use an external clock and  the advantage may be levelled out.

Regards position of the dexa clock, crystals are hugely ambient and electrical noise sensitive. As an example whilst trying to tune two DIY built clocks to run at the same speed at either end of the USB link the clock speed would go haywire just putting a hand or screw driver near a to a VCXO module that I was trimming on the 'scope the wave form also became slight less stable.

I finally got round to implementing proper shielding of my audio and 2 USB Dexas clocks today. There are some new components in the clocks now that need to run for a few days, but even with this the difference in performance already is massive. I just cannot believe what has happened  its given about half the benefit of upgraded clocks again .... I'll post on this properly soon.

Nick.

I should have read the post more carefully ! I had not noticed from the above Peter that you have the NEC version.

It might be fun to try the TI version still but NEC copy of the card first to try out :-)

Nick.