@tedsmith So tell us about the new Mk II. A new xylinx, new Power supply, output transformers, no display, isolated inputs that drop power? You have been sprinkling us with hints. Any new developments
Why would one opt in early versus waiting for the TSS?
PS Audio is picking the display, I donât know where they are headed on that one.
? Unused inputs can be powered off, e.g. USB, I2S1, I2S2, ⌠Is that what you are referring to?
If the budget allows each input and output will be isolated and have separate controllable ground lifts.
If budget allows there will be some output configuration options, e.g. optional 4VRMS balanced on the RCA output, optionally lift the ground on the output transformer center tapâŚ
If budget allows the XS4400 (unless I find a better one with a lower cost. 
)
2 AES3 inputs, optimally to be used in parallel for quadrate.
Instead of simply a high and low output level, the outputs will nominally be standard (4VRMS for balanced and 2VRMS for unbalanced) and then there will be a series of options, say, -6, -12, -18, -24⌠or -10, -20, -30. More options cost more money so weâll see where things land. They wonât be output attenuators so there shouldnât be any sound quality change as they are changed.
Minor DS bug fixes (the FPGA PLL power supply on the DS wasnât isolated from the output supplies, oops.) The I2S 5V outputs will be reverse current protected so external I2S devices canât try to power the DAC.
Some simple new convenience things like trigger outputs to power up amps, etc.
Budget? FOMO?
Any thoughts about the possibility to use existing PWD/DS case when upgrading to DS Mk II?
Ted, when say âthe FPGA PLL power supply on the DS wasnât isolated from the output suppliesâ, are you referring to the 12V analog board power supply?
Ted,
Thanks for the reply
Yes it was.
Minor DS bug fixes (the FPGA PLL power supply on the DS wasnât isolated from the output supplies, oops.)
Isolating the supplies is this something that would be equivalent to the boost one may expect with implementation of the external supply that users are trying in this mod thread?
Thanks, @wiesiek! that helps. Definitely Iâm going with the XS4400: it is the safest option 
Hi @vee
Your solder work is absolutely fantastic (Iâm not very handy with the solder iron, but itâs clear that you are).
What was the used solder?
In fact, for everyone that already did the upgrade, which solder do you use?
WBT-0800 Silver solder, fine silver content 4%
I mostly use Johnson IA423 solder (shown on XS4400 pins), which is eutectic and 4% silver, I believe. Excuse the apparent messy-ness of my DS, but shows some shielding and Belleson regulator (on 5v input board and 12v supply).
On the first day of burn-in it sounds great, with more midrange presence and tighter bass, but a little less sparkle and decay. Hopefully the latter will improve with time.
No. The top level power supplies on the Mk II digital card and the analog card will probably be of a similar design to the TSS. They will take their current in phase with the voltage and will take it evenly during most of the cycle - they will have a high power factor. Most linear supplies fill their caps near the top of the voltage peaks in the input power, they chop off the tops of the voltage waveform and take big gulps of power near that peak. The new supplies in the Mk II will take most of the current away from the peak. This will have two effects: the Mk II supplies will stay away from the dirtiest part of the AC power that most audio equipment uses and their effects (if any) will be away from the part of the power that most of the rest of you system uses.
The power trees on the digital card and the analog card are different but share the goal of keeping the power supply noise away from the main part of the analog card. They both regulate to a rough initial voltage, around 14.26V and 7.12V on the analog card and 5.47V on the digital card. The digital card then regulates that to 5V with a quality, but pedestrian regulator. Most of the digital card runs off of 3.3V so most of the 5V current is further regulated to 3.3V. The FPGA also needs a power supply for the PLL at 1.8V and the core at 1.0V. The 1.8V comes from the 3.3V and the 1.0V comes from the 1.8V.
There are also isolated power supplies for each of USB, I2S1 and I2S2. These supplies use a isolated 5VDC to 5VDC regulator with a built-in 5kV isolation barrier. Then the USB voltage is further regulated to 3.3V and 1.0V for the XMOS (USB) chip. The I2S 5V gets regulated to 3.3V for the I2S receiver and the digital isolators. The I2S 5V also has a âideal diodeâ to supply the I2S connector with 5V.
On the analog card there are multiple high quality regulators: each of which is better than the myriad of audiophile regulators out there: they operate over a wider bandwidth (up to 10âs of MHz) and have very large input and output capacitances, etc. The 14.26V is regulated to 12V then each channel of audio regulates this to a solid local 10V for the output rails. The 7.12V is regulated to 5V for everything but the output rails: multiple separate regulators take the 5V to separate 3.3Vs for the clock, the digital reclockers, the misc CMOS, the isolation relays, etc.
The control processor on the digital card can measure each of the 8 analog regulator outputs and six of the digital regulator outputs. We. skipped measuring the USB and I2S regulated outputs since that would involve getting an quality analog voltage measurement across the isolation barriers.
The stuff above red scribbles is just the top level 14.26V and 7.12V regulation (the stuff to the left of the scribble is the 120V AC stuff):
The other six regulators are distributed where needed.
On the digital card the top level 5.47V is inside the red scribble:
(You can see the via stitched isolation barriers on the digital card and the output grounding options on the analog card.)
It looks like a reasonable audio transformer, but too good of transformers arenât right for the DS. For the DS a higher leakage inductance is used to help filtering.
Hereâs a typical (but simplified) linear power supply: (component values were chosen to magnify the current gulp):
Here are the simulated output voltages and currents and power draw, you can see the current gulps:
Hereâs the simulated voltage, current and power draw of one of the DS Mk II supplies:
Despite the weird shape of the current draw, itâs much âsmootherâ than a typical linear supply, it draws current/power much more evenly.
Exciting!
Thanks for the power tutorial. The above statement makes the Mk Ii DS of interest and at least worth the 30 day trial vs the DIY hot rodded DS.
How well would a P20 regenerator address the current gulps at the dirty peaks? It seems a regenerator tries at least partly to address the inferior Big gulp supplies unless they all compound gulping in unison based on music or regulation frequency?
A smooth pull of current might also make the DS Mk IIless sensitive to power cords correct?
In terms of stepping up the ladder to the level of other high-end peers, this dual input strikes me as a must have.
Glad to hear it is on the drawing boardâŚ
FWIW.
Thatâs why the P20 (and all PS Audio power regenerators) have a low output impedance - that basically means they have the extra energy available for when higher current is needed. Another way to look at it is to look at the before and after voltage pictures on the P20, the output voltage is pretty clean so the regenerator is keeping the voltage correct in spite of the uneven current draw.
Yes, a design goal of the DS, DS Jr, TSS and DS Mk II is to draw current as evenly as possible lessening current related noise on the power lines. The newer supplies also try to shield the outside from noise in the boxes - thereâs a lot of storage capacitance and explicit high frequency regulation to keep the outside noise out and the inside noise in. Iâm sure that the Mk II (and TSS) will be less sensitive to power cords, but the more reveling a system is the easier it is to hear minor things like cord differences.
Ted⌠any DS MkII release date?
Thatâs up to PS Audio, I believe the approach will be âwhen itâs readyâ. They have a lot of software to write, getting parts and building boards takes time, designing the box, ⌠In some sense these boards and the FPGA software are just the beginning. Paulâs post mentioned something but I donât remember exactly what.