Why does dcs prefer external clock and internal power supply

i have a rossini and a clock.
i was just curious and wondering why do dcs designs prefer an external clock and a internal power supply

  • in simple terms, it can be argued that an external linear power supply can reduce a lot of noise if you move away a giant transformer far away from those digital circuits
  • all high end products like ch precision, pass labs , simaudio , nagra etc have an external power supply strategy for their digital front ends and some for preamp and power amps too
  • coming to an internal clock while i have no idea about this but a customer definitely can save on 3 high end cables ( 2 digital and 1 power cable )
  • on a serious note what cud be the inhibiting factor to build a clock outside but a. power supply inside?
  • and dcs is anyways about multi box strategy on their lines then why not think about an external power supply where conventionally so much gains in SQ and noise reduction are made.

would love to hear some thoughts around this.

1 Like

The master clock in a full dCS system , for example the four box Vivaldi, has to used as the same timing source for all of the components or the entire purpose is lost. Therefore, as the master clock is not unique to any of the components, it is rational for it to be a separate item.

I am sure that a marginal improvement in SQ could be found using external power supplies but the added cost would be considerable in just the additional chassis alone ( you would not believe how much just the front panel of a Vivaldi component costs dCS) and that the amount of real estate in racking terms needed to house an 8 box system plus the two box amplification components from other manufacturers would, no doubt, limit the size of any potential market.

There is also an historic view. dCS was originally a studio equipment manufacturer with no products for home use. Eventually some enthusiasts started buying that studio equipment for home use. dCS realised that there was a potential home market and, for generation 1 of the home systems, they basically just put the circuitry intended for studios into nicer cases for domestic consumption. So the system architecture of the studio gear was inherited by the domestic range. This has more or less continued not least in order to ensure a degree of backwards compatibility. This enables users to upgrade piece by piece and/or mix components from the ranges rather than need the significant amount of capital to buy e.g. a full Vivaldi stack in one hit.

However a new generation of dCS products will, no doubt, appear in due course so who knows what that may bring.

@PAR thank you for beautifully articulating the historical reasons for the same. appreciate it

  • i agree with your rationale for the clock
  • but i also think external PSU offers substantial SQ improvements rather than marginal and the proof is pretty much available across top segment of high end audio manufacturers
  • also i am more than happy to pay 9k for. a power supply rather than a odin or magnum opus cable.
  • i feel for Rossini segment and price point, a built in high quality clock makes a lot more sense
  • for vivaldi owners who want nothing but the best , they should absolutely demand both an external power supply and an external clock no doubt.
  • power supply units from nagra , ch precision , simaudio they all power 2 or 3 units
  • so one power supply add on to power a full vivaldi stack should be easily feasible and its not rocket science at all and i think vivaldi owners should demand it
  • even Rossini segment future products should get an option for an external power supply.

its conventionally wise and there is not a single high end reference grade product that has not benefited from an external power supply and its a no brainer IMO

hopefully we see this in future which is both a value add and a solid foundation for great products.

1 Like

Well take naim for example, they love a power supply box, and they are more than happy for owners of the Nd555 to run 2 x 555 power supplies, even if said power supplies are now only using half whats in the box, and even then with running 2 power supplies its cant match what a rossini and clock can do.
I guess it all has a lot to do with how the power supply is made and how they go about this, naim once again like to use a big transformer and regulators, this i am guessing is a noisy way to do it these days, probably good for power amps, but a digital box, just pulling very small loads at low voltage, probably not. Plus naim only have a few box sizes, and so would rather put that into 2 boxes that they have, rather than have to make different sizes, plus then they can play the power supply upgrade gain, win, win really and a great market strategy.
Others may do the same, i dont really know, but for me i like the way DCS go about things, and the only thing that would make the rossini better, is if they had put the upgrade clock in it as standard, but then like naim i can see why not, probably.

1 Like

@Dunc i agree power supply need not be big boxes with giant transformers at all.
take for example lumin x1 linear supply or sbooster or paul hynes PSU they are small little outstanding psu to feed digital equipment with low current requirements.
my point is externalizing the power supply is well known and proven strategy and always has tangible benefits.digital equipment environment are even more sensitive than analog , so all the more makes sense to externalize the power supply.
that is the reason nagra and ch precision do it and i do not think it’s a marketing ploy.
reference equipment like vivaldi should extract every ounce of performance and i think external psu can be a game changer.
they can easily add one small/medium size housing/psu that can power all 3-4 vivaldi stack units.

reg rossini clock atleast they have given an option to add a external clock which increases the performance.
in the same way there should be an option to add an external power supply as well.

To me, it’s all about the execution. If the internal PSUs are competent, I am fine with that. My MSB has an external PSU, my Vivaldis do not; both sound superb. But I’m game if an external PSU could contribute to any or all of: (1) audibly improved performance; (2) measurably improved noise reduction; (3) even more attractive main component case design; and/or (4) reduced per-component cost (stop laughing!).

I know everyone likes to think an external PSU could automatically make everything immediately better. I don’t think that mere externalization alone is inherently superior. It has to be done better to be better. And I think one would be hard-pressed to prove that externalizing the PSUs of a Vivaldi stack would inevitably make it better. The stack has been favorably compared to many digital front ends with outboard PSUs, and it has not been found wanting. And this has been true year after year. My bias is to believe that dCS would make it better, but that doesn’t mean we would hear the difference. Game-changer? How so? SQ? There is no real evidence to support such a belief. Market competitiveness? Maybe, but I don’t get the sense that dCS has been unable to sell due to lack of external PSUs. Cost? Doubtful.

Also, as suggested, it would need to be one unit for a full stack. Separate PSUs for each component in a full stack would not be welcome chez here. :wink:


@PaleRider -

  • i think externalizing power supply always has benefits
  • i wish all hifi equipment ran out of DC
  • removing the emf radiating transformer and putting them out away from sensitive audio circuits seems to be logical for me and especially for top of the line reference gear
  • also the biggest proof of this is in the vivaldi stack vs vivaldi one.
  • everybody says vivaldi stack sounds better than vivaldi one.why its the same thing and same boards crammed into one case.i think it sounds better bcoz of isolation provided by the power supplies.

Hi there Greg…filling a bit of time before we open the prosecco on New Years Eve! :clinking_glasses:

I searched for, ‘power supplies for dCS products’, and came across your post.

I have Naim amplification. According to Naim, and their community site, a separate power supply is a must on their amps and almost biblical on their digital source.

Apart from a Clock, which I’m saving for, to complement my Bartók, dCS don’t appear to suggest having a separate Power Supply for their products.

I’m thinking that dCS feel that their own power supply is more than adequate?

Best wishes for 2021

David :+1:


Naim, but YBA as well, Moon and Pass Labs to a lesser extent…I guess there are many other manufacturers doing it…

David, I have not noticed any negative aspects of dCS performance that could be laid at the door of them having inadequate power supplies neither,I expect, have you. So I think this is basically a rhetorical question.

Of course big power supplies that need a separate box may indeed be appropriate. Perhaps even more appropriate in an analogue amplification scenario.

The original Naim pre amp had no power supply of its own and was powered from the power amplifier. The first separate power supply, SNAPS came later. At the time the Naim 12S ( or K) plus the 160 or 250 were amongst the most costly solid state amps made in the UK. Julian Vereker was a canny entrepreneur and I would not be surprised if powering the preamp from the power amp was not a way of attracting customers by keeping the cost of membership of the Naim club lower than would have otherwise been the case. Of course once SNAPS came out and was successful this provided an upgrade path. JV and Ivor Tiefenbrun, in their mutual axis of the time, were keen of getting customers back in the shop for an upgrade on a regular basis. Selectable power supplies is a great way of creating repeat business. Cynical? Not really, a company has to make a profit to survive.

Of course not all two box components divide the responsibilities between active circuitry and PS. Some have a “clean box” , “dirty box” configuration.

I am sure that dCS ,in their 32 odd years, will have considered these approaches but, at the time, rejected them for various reasons just as I expect they have considered and rejected adding room correction dsp or a tube output stage and many other options. Still, who knows what may come along in the future?


Agreed, Pete — it’s not as if dCS are shy about box count. I suspect they’d have done it already if it helped enough.

Mind you…if a full-stack Vivaldi replacement in 2021 (or whenever) has eight boxes, four of which are power supplies, we’ll have at least part of an answer. We’ll need to work out whether it was done for sound quality, dCS bank account quality, or a combination, though :stuck_out_tongue:


Well, if so, it won’t be coming here :grin:.

I would prefer two box solution like vivaldi one but slightly different

dac, upsampler and clock into one box
one power supply with 3 or more DC ouputs to feed the dac, upsampler and clock modules
would be great if they can come up with modular approach too for upgrading

iam not so sure that dcs would have evaluated and ruled it out.if they ruled it out then my guess is its cost or pricing based decision.
good external dedicated power supplies cost a lot of money .
say for example CH precision X1 costs 17k USD

As someone who has run a Naim 500 system for almost 20 years I can say that overtime many Naim users start to suffer from box fatigue. Over the last number of years I have made an active attempt to reduce the box count by doing such things as moving from a Superline and Supercap to a Linn Urika phono stage (1 box saved), moving from a Naim CD player and power supply to a Linn Klimax DS (1 box saved) etc.

As someone new to DCS this year with a Rossini and External clock I can say that a full Vivaldi system does not interest me simply based on the number of boxes and cables. On the other hand the older Vivaldi player one box solution remains extremely interesting.

I have many Naim friends who recently have abandoned the brand and gone with a super one box integrated such as Vitus, Accuphase etc. In fact there is a thread on the Naim forum with people asking for a high-end top of the range single box super integrated solution.

Imho as the current generation of audiophiles get older many start to seek out simplicity. High box count, extensive cabling, and multi tier rack systems no longer fit in.



Just so Pete. dCS seems to have little concern for box count. I suspect they found little benefit to a separate power component. I think the whole “separate PSU” trend in audio has become a bit of a fetish (says the guy with separate PSUs for his Wavelet, Townshend Allegri Ref, SRS Perf10, etc.). My MSB DAC has a separate PSU, and the Vivaldi does not. They sound different, but I suspect the aural differences have much more to do with the differences between the RingDAC and the R2R. I have gotten to the place where I don’t see it as a prerequisite to stellar performance.

Ultimately, whether to design a component to use an integrated or outboard PSU should be driven by objective measures such as the noise floor of the component - if you look at Stereophile’s DAC measurements over the years, probably 4 out of 5 of the best measured (lowest noise floor) DACs have integrated PSUs.

So it would seem that insisting a component has to have an outboard PSU makes little sense when it comes to actual performance results. :slightly_smiling_face:


:+1: :beers:

Engineers evidently have their own opinions and preferences. John Siau’s designs for Benchmark have very low levels of noise (see Stereophile’s measurements or the Benchmark manuals). He claims that to get the same level of noise that he achieves with a switched mode power supply (in the box with the amp or DAC electronics) a linear supply would have to be in a separate chassis a few feet away. In fact - and here is a good starting point for a long and passionately argued thread - he says “in my opinion they (linear power supplies) have no place in true high-end audio equipment.” Of course, they are almost universal…


Simon, I have heard more than one designer/engineer say that. As I said, for audiophoolish consumers, separately housed LPSUs have become almost a fetish, something one dictionary defines as:

an object regarded with awe as being the embodiment or habitation of a potent spirit or as having magical potency

Sounds all too familiar.


Mind you, dCS use both linear mode and switch-mode amplification.

Anyway in the Bartók headphone DAC:

Amplification for headphones

In the past, it has been common for dCS owners to use the existing line out stage on their unit to drive headphones, by making up a special cable to run from the outputs to the headphones. It actually does this job reasonably well – even though not designed for the purpose – with its low output impedance and low harmonic content and distortion, and its ability to drive a fair amount of current. When we designed the Bartók Headphone DAC, however, we decided to add a specially-designed headphone amplifier to drive a wider range of headphones (particularly low-impedance types, which wouldn’t do so well on the existing line output), both balanced and unbalanced, directly, in addition to the standard line outputs. The design, which attempts to retain the essential character of the existing line output stage while adding the functionality described above, follows a dual-mono configuration, keeping the two channels completely separate. The amplifiers are powered by a combination of linear and switch-mode power supplies that are completely separate from those employed in other parts of the product, particularly necessary as quite high ground return currents are experienced with low-impedance headphones and these need to be kept
well clear of the DAC and associated circuitry. A substantial toroidal transformer is employed in the linear PSU.

From: [page 11]:

1 Like