Sigh…That’s not what he means by “long term stability”…
Remember these answers are for my DAC designs, not necessarily all gear:
Long term voltage accuracy isn’t terribly important since there are multiple layers of high quality regulators closer to the sensitive points to handle that. (I say “long term voltage accuracy” because short term voltage accuracy is basically noise.)
Lower noise is always good. The filter, reservoir and bypass caps and lower level regulators will filter most noise, but not getting the noise in the first place is clearly better.
This is clearly important, but it isn’t often messed up unless someone changes components to something not expected in the design (for example to caps with too little ESR or vastly different capacitances.
Two different answers depending on which way “capacity” is intended:
Loosely speaking there’s no such thing as too much current capacity. But too much is wasteful both in original parts costs and long term running costs. However it is the case that some power supplies are designed expecting a certain minimum current draw.
If instead you were talking about changing caps in a power supply to larger values, the first caps after the bridge can be too big, if so 1) it can take a longer time for the supply to reach it’s nominal level after powering on or 2) bigger caps can draw higher current (for a shorter time) during filling on each cycle of AC. Drawing higher current might not be good and drawing higher current for a shorter time makes more noise.
Lowering EMI is always important. It may be more important to other things in a user’s system than to the DAC.
The DS and TSS draw current at a pretty steady rate, so if the average current available is high enough then the supply impedance doesn’t matter as much. (There’s plenty of extra storage capacitance on each board.)
5 Likes
Ted, does this mean your old or your new budget was equivalent to the 20k sales price forecast?
mumetal is for magnetic shielding. Any conductor can be used for electrical shielding (e.g. a Faraday cage.) Some are more efficient than others.
3 Likes
The first design was for a higher price point. All of the public discussions about cost and price have been about the current design.
1 Like
1cm? That’s like 0.4 inches. I’m not a shielding expert but to shield you need more than a plate. You need to completely enclose the area to protect and if there are any holes they need to be smaller than approx. the wavelength of the frequencies you want to attenuate.
1 Like
If you have some ERS sheets on hand it’s worth a try - but remember they are conductive so be careful.
1 Like
Done some “heavy duty” modding to some equipment in the past, most recently pretty much a complete overhaul on my Oppo 203…
but I’ll sit this one out and wait for the next Mountain to drop. 
1 Like
Wow, so Paul initially wanted to offer a 40k DAC…
While I appreciate the hyperbole, I am comfortable this is not true.
1 Like
ok but then why was the previous budget double?
I wouldn’t say that as much as they reconsidered the budget when it came time to build the prototypes. Since PS Audio’s pricing is a simple multiple of cost and the price was unexpectedly high, the cost had to go down. I wasn’t happy at the time, but I’m much happier with the current design. For example, the original design would have required 2 multiple winding transformers for the digital and 4 for the analog. Now I get the power I need with half that number. Since the power supplies take a significant part of the boards, the difference in cost for just that change is significant.
Here is the original 3D view of the analog:
Here’s the current analog:
Old digital (After ripping off one of the power supplies):
Current Digital:
4 Likes
Very interesting! I think it must be great to experience the ways to achieve similar results with less budget…a hell of know how you shouldn’t immediately publish in detail I guess 
How do you see your competitiveness in terms of performance/price point to other digital manufacturers who may have different calculation strategies like putting in development costs vs. pure parts by factor?
Is it better for you like that? I guess yes given the time and energy you spend with development.
I think PS Audio customers get a bargain just considering the parts cost. One could fairly ask if other designs could get by with cheaper parts, but then I wouldn’t be building it.
In reality I won’t ever get my development time paid for, but I don’t do this just for money.
16 Likes
Yeah that’s what I always understood…we honor it! 
5 Likes
Unfortunately the reality is as always everything is built to a price. even the MSB. they could have made it a million dollars i am sure. then maybe 5 people would have bought it. I personally would have liked to see it have been $40K. then I would be comfortable that with Ted behind the wheel it would be the finest DAC on the planet. It just might still be. we shall see. I have the 30 days. With the longer cord I know I can now actually make good use of it. that at least is good news.
I do not program just for money either. Although I have lost a company 3 Million dollars in 6 hours before 
I bet Ted has had some mishaps too? Maybe not.
Ted, Speed racer claimed I am a fool. Is that not what you meant by long term stability? Did you mean for how long it is powered on rather? Since Lavry specifically states stability as it ages. I am also/was a pro musician. I was just wondering if you could please clarify what you meant please?
Speed-racer does not care for me very much for some reason. I can actually think of a lot of reasons not to so I am okay with that. Plenty here are okay with me and that is more than I ask for. Especially the fact that I have yet to alienate you, Ted and Paul. I am glad for that. I know I can come off as a jerk.
adriaan, try to obtain a Hynes PSU. Probably going to get it used so the price is a lot still but not horrible. It is the best ever made but no longer made. It far surpasses anything you can get now by a mile. Finding the right voltage if it is non standard is unlikely though but a competent shop can easily fix that for you. You will also need a custom adapter for your device made. Again, any good audio repair shop can do this.
I did not want to open yet another useless thread here. so I will just say this here. I wasted money on jplay in my feeling. My URC TDC-9100/ MRX-30 do a considerably better job IMO. They have two way communication. The 9100 is a 10 inch tablet. It shows all song info, cover art ETC, browse your library and much more than Jplay. I do not need anything to make my system “sound better”. I am running “core windows” on a purpose built NUC with Audiophile Optimizer in full ultimate purist and fidelizer via Kernal streaming from Jriver in DOP. I have a complete outboard signal chain before reaching the DSS. Galvanically isolated, relocked both Femto and ocxo,buffered, ground removed…All devices powered Via Hynes LPS’s. everything into a PS Audio PP-20. All AC-12 cords. I use multiple sections of Crystal cable Future dream USB cables. Luckily Ted figured out i had too long a run. Hat’s off to Ted as usual! I was just mentioning I personally had no use for jplay but I can see where others would.
I just want to say the DSS with the Bridge was perhaps the best $6.5K I ever spent .Being upgraded almost 7 years later! Try that with a cell phone! The sound honestly nearly approaches the MSB and even 069 CD while streaming Flac radio! Simply amazing. I have a feeling that the next update will cause it to actually surpass the MSB! Of course not on local .WAV files though. Still, there is a bit of a price difference. I love the DSS.
I love all of my PSA products and love the company. I will support them for the rest of my life. My Cullen GCHA to this day I cannot find but very few Headamps close to as good. I have one but is is 20x the price. I just know of no other company anything like PS Audio. Talking to the engineer and CEO in the middle of the night on a weekend. Wow. Yes, I am a fanboy. The only thing, I would like to purchase Pauls’ IRS speakers. I stupidly sold mine. If he would really sell them I can’t really get them because of the logistics involved though. I wish. Anyways you guy’s rock.
One last thing. I failed to mention that I have made one internal modification to my DSS a long time ago. I sparingly used ERS paper. That is all.
Keep up the good work Folks!
Any streamer will be a source of interference because (among other things) they have to have a processor to deal with Ethernet, USB and/or the streaming protocols and things like file decompression, etc. The best way to separate it is to keep it external to the DAC. I.e. I would use an external streamer of reasonable quality and I2S (for things like double rate DSD) or S/PDIF, AES3 or TOSLink. I would also remove the now unused Bridge from the DAC.
3 Likes
Stability of a crystal (or oscillator in general) is how well they keep time in spite of, vibrations, temperature, time, etc.
Short term stability is essentially jitter/phase noise. Long term stability is accuracy over macroscopic time, temp, etc., often measured in parts per million per year, ppm over the life of the crystal, ppm per degree C, … It’s semiobvious that controlling the temperature (with an oven, etc) will help control the long term frequency drift. Also some oscillators have mechanical isolation to keep vibrations from affecting stability.
1 Like
I’ve done several mods to my DS - usually in a way that can be reversed if the results were not positive - and in my experience it responds well to power supply upgrades and RFI shielding.
I’d like to try an output transformer swap at some point too, such as the Jensen Ted is using for the TSS, but I’m a little nervous because the extended high frequency performance will change the low pass filter characteristics and might require additional filtering. Conversely, I understand the FPGA has gone from originally converting to double-rate DSD to quad-DSD now, so maybe additional filtering wouldn’t be necessary. Maybe Ted can enlighten us here. 
1 Like