Recent S5 update

It is almost a year and a half since we started shipping the S5, and, as it was announced on the main S5 page, this DAC recently got some updates. More information about them follows below, and some of the information and photographs of the S5 internals you might actually find premiered here.

First, the main improvements in the audio part are associated with the output stage supply. The previous year, when I was designing the AYA 5 (at that time the S5 was already released), I was asked to redesign its output stage. I did this, and I redesigned its supply regulators too. The result was very nice, and since some of these solutions were used for the first time in any Audial DAC, I applied them in the S5 too. Of course, with respect to the AYA 5, the S5 still brings the advantage of separate supplies for two channels, and the S5 also employs better passive filtration.

Technically speaking, the S5 output stage regulators now have a lower output impedance. Sonically, this change brought a bit more definition and refinement, and a tonally more correct presentation.

And, let me publicly address the questions on possible upgrades I am receiving these days from previous S5 owners. Although the difference is audible, and “retrofitting” is generally possible, and although the circuit changes are not that complicated, due to a major use of the SMD parts, it involves a certain “surgery”, and in that way, it is not very practical. Also, strictly speaking, a complete upgrade would also require increasing the supply voltage, which would then require a somewhat different mains voltage transformer. But even putting that part aside (as less important), and assuming only the circuit changes, at this moment I am not very prone to do this if that would require shipping across the border.

So, here is the picture, showing the left channel output stage and its supply. (The picture will look better if you click on it and see it at higher res.) The right channel looks, of course, the same. Previous S and AYA DACs owners will also notice the missing polystyrene capacitors, which bypass the I/V resistors. These caps are now replaced with polyphenylene sulfide (PPS) parts, mounted at the bottom board side.

And, it might be boring to repeat this, but my experience suggests that I still have to stress: the S5 output stage, which is only a slight evolution of the topology introduced with the Model S (2008), uses no opamps. It is the transimpedance stage, with an output voltage buffer, and it is entirely “feedback-free”.

See also:
S5 2024 updates (S5b)

7 thoughts on “Recent S5 update”

  1. And a few words on the S5 USB stage.

    Still being very similar to the previous Audial USB topology based on the XMOS processor, the S5 is the first and still the only Audial DAC employing a completely linear supply for the USB decoding part. (I am not sure if it is also the world’s only, but I am sure there are not many other such DACs, if any.) Consequently, due to about 300 mA current required here, its pre-regulator series pass transistor uses the substantial heatsink, which again, due to the specific environment, had to be custom machined (and anodized). A 1V regulator also employs its own small machined copper heatsink, attached to the thermally coupled local copper plane beneath the board.

    Although some parts are hot to the touch, all the temperatures here are well within the specifications of all the parts used (accordingly, all the SMD capacitors here are either NP0 or at least X7R). One year later, I can only say that the whole setup appears to work reliably.

  2. Hi Pedja,

    I have seen today on Head-Fi forum a new owner presenting his S5 in black, it is looking GREAT.

    Please, can you share some info about the black version?
    What about availability and price?



  3. Thanks, Matt.

    The black chassis is made only on request. There is no guarantee it can be made every time, but practically speaking, requests were mostly positively resolved.

    The black finish does not change the price, but the lead time is normally longer, usually 30-40 days.


  4. Páll Silvan Ennigarð

    Hi Pedja
    Can I ask about heat? Can the Dac be left on 24hours 7 days a week? Is there a standby mode, have asked before, but is a remote planned, if not, is it because sound deterioration?

    PS Ennigarð

  5. Hello Pall,

    I do not recommend leaving the S5, or any other TDA1541(A) based DAC, on 24-7. This chip is the old TTL architecture, and it will not last very long if left constantly on (its DEM circuit usually fails first, probably after a couple of weeks).

    As for the remote operation, the S5 does not support it for now, but as we have just released the remote control unit, it may be added at one point. Since the command logic uses a separate supply, and we design controllers to enter sleep mode when they do not perform actual operations, it will not introduce any sonic performance compromises. There is no time frame for this for now, though.

    Also, the remote control probably would not provide a stand-by function for S5. In my view, such a feature, if done properly, demands a separate small transformer and supply driving the powerful switching relay for the main transformer, and all this would then take its own space inside the chassis. Mind you, the existing remote control unit does not include the buttons for this function, either.


  6. The AYA 5 and S5 are very similar in many ways, especially regarding the overall topology.

    The differences are, speaking technically and most shortly, that the S5 has a more advanced supply, and a more advanced output stage. Also, the input switching circuit is completely different (this is probably obvious).

    As for the consequential subjective and objective performance differences, you might want to check my posts from October 3rd 2021, on the AYA 5 topic.

Leave a Comment


Scroll to Top