So yes, there finally is one :
Ultra low power supply everywhere (USB Interface, Clock Section, DAC, Gain Stage).
New audio oscillators.
100% isolation of USB Interface and DAC with zero added jitter by the isolation by means of game changing "invention".
Net peak-peak jitter at the input of the D/A chips went down from 210ps to (way) less than 13ps (13ps can be regarded the limit of the analyser used). In RMS jitter this will be (way) less than 1.625ps. See the "jitter" topic for this.
No influence of the PC any more. This also means no XXHighEnd dials influencing any more.
Better DC Offset Control because of the elimination of one formerly separate ground forming point (now only one remains).
A total of 4 newly added PCBs will do all this.
What will happen to your NOS1 ?The power supply in the left leg will be removed and replaced by an ultra low noise regulator with measured positive rail supply noise of ~2uV.
The power supply to the actual DAC board will be through an ultra low noise regulator of ~1.5uV for both the positive and negative rail.
The power supply to the Gain Stage will be through a separate ultra low noise regulator of ~1.5uV for both the positive and negative rail (no feedback from the DAC board's supply possible).
The DAC board will receive another PCB on top of it that from now on being the "clock section". The both oscillators on it will receive their supply from yet another ultra low noise regulator of ~1.4uV (positive rail) with no back or forth feedback to the DAC board.
The both oscillators (one for 44.1KHz base and one for 48KHz base) will mutually be powerd down when the "other base" is in order. Thus also no feedback possible from the respective other oscillator.
To shut down the respective oscillator, a new version (1.04) of the Phasure NOS1 Driver is required when your NOS1 is used outside of XXHighEnd.
The output voltage and maximum current possibilities of the main (shunt) PSU will be changed to the now required needs.
New (more) wiring will emerge between the USB Interface and DAC board.
The oscillators will move from the USB board to the new top DAC board (and will change type along the way).
The output terminals will move position to the left (seen from the back side).
The holes now emerging will be nicely closed.
The CPLD ("FPGA") on the DAC board will receive a new program.
There will be an "A" signature on the front in between the legs.
There is more detail hardly worth mentioning because you are not interested - like input capacitors needing to move from the one PCB to the other or other wiring and connectors needing change. Still, that too consumes time to perform.
Anyway envision that your complete NOS1 needs to be taken apart for this in the first place and that it has to be re-assembled afterwards. All of it (ok, not the display).
Functional MeritsFirst off, the crazy thing : While XXHighEnd was made to influence the DAC for the better, this "software playback" influence is now out of the equation. Of course, right from the start (at designing D/A converters) I have been trying this, and now it finally works out.
Although you don't know it, two very different kind of DACs have been designed in the past two years (up and running) with the same objective, but THIS is now about how you NOS1 owners can upgrade to that situation. The approach is very different now (and the most new) and I guess it could only emerge from the experience of the past two years.
Again, you don't know a thing about this ...
While software can not influence any more, the PC itself can't either. In the end this is logic (at least to me), but notice that things like a Silverstone USB3 card also lost their influence. Or USB2 vs USB3 for that matter. Or BIOS settings. Devices shut off. Name it.
There is just no electrical connection any more between the PC and the DAC;
Everybody will say easily "yes of course, because isolation is isolation, right ?". Well, true in itself, but if it were that simple it would have been in the first NOS1 to begin with. And don't you think I haven't been working on that right from the start (like about Adnaco glass interfaces which (as a few know by now) also fail. Thus, 100% isolated for sure, but *such* isolation only makes it worse.
Side note : I am pretty confident that for the NOS1a such a glass interface can be applied now, hence it would allow you to have the NOS1a on quite some distance from the PC and nothing will deteriorate.
Without having said it explicitly, it will be clear to you that no USB interface "tweaks" etc., or USB cables for that matter, will influence one tiny bit from now on.
Done with the tweaking - just good sound.
Ehm, THE BEST SOUND.
Oh yes, it matters. Say that it can matter to the degree you are used to from one setting (or tweak) to the other, but now a much bigger step and to the better direction only;
As often I don't have words for it, so I hope you will just trust me. But it is better-better-better.
This "better" has been apporached from all angles possible, mainly thinking from my ever so often brought forward "noise" subject and how less noise can imply less good SQ because of now remaining more distinct frequencies. Well, for those who can make out merits of supply noise : that is now way UNDER the output noise of the DAC. So the 8uV (or something like that) measured at the output is not related much to the "incoming" supply noise, however, the supply noise is detrimental to jitter at various places (but think oscillators to begin with). Now, supply noise previously superseded the DAC's output noise by far, while now it is 1.5uV (RMS) only (think around -160dB).
Notice that two years ago this would not have been possible to achieve because of new components (think chips) available now.
For insiders : For minus supply voltage to go this low no components are available, so this was self-created.
So what does this by now mil-spec machine bring us ?
For the obvious part : I can run my processor at any speed I like, I can leave on the LAN, I can play Attendedly - I can do anything to my convenience.
And the expected part : SQ is increased by so many miles that it is hard to bring across. More (SPL) highs are pushed out than ever before while at the same time totally nothing is hurting. Somehow it is all about balance and don't ask me how it is possible that e.g. 70's music had covered up so much good highs like better than today's (without replaying today's for comparison). Things changed into creapingly real and the difference between any drumming heard and our son's practicing has been diminished to undisturbing (really so !). The real "un-difference" still comes from my self-recording of the drumming, but this is because I can compare it back-to-back.
For the past couple of months I had to lie somewhat or keep quiet where I said something anyway, but I had this running in lab-setup for these couple of months with all sorts of trial PCBs and wires to plug all together, while now all is running from the meant final PCBs and ... working. This latter part was related to the thrill of lab setups working while the real thing for example was about the shut off of respective oscillators which could not be simulated. So now you also know why my sampling rate was fixed to one rate only and for those who got that : how I had to break-in a new DAC because I needed the old one for XXHighEnd measurements etc. Meanwhile there was no way to revert to my old NOS1 for listening because the difference with this "NOS1a" is just too huge.
What's next ?
Peter
