Auralic Leo GX reference master clock

Audio Systems
1

Does a Auralic Leo GX refernece master clock also works on dCS gear?

2

No, I think not. The frequencies are 90.xxx and 98.xxx MHz, the connector is screw-in rather than the usual BNC push and twist, and frequency switching appears to be controlled via another connector. So, incompatible all round! And in an Auralic system it seems to replace rather than supplement the dac clock (the dCS approach).

I understand your interest though: much cheaper than dCS.

3

It also takes an hour to warm up before you can use it, I owned one as part of a full G2 stack, which was replaced by a Bartok and I never looked back. A good mate made the same move, heā€™s now Vivaldi and Iā€™m Rossini.

The Leo was the least impressive part of the whole stack, same mate had same opinion. The Aries however was impressive, the entire Aries product line has been very good since day 1 IME.

4

Yes, dCS products use the following protocol:

  • Word clock (I.e. the frequency is the sampling rate of the file or an exact integer multiple thereof)
  • Square wave (usually called ā€™TTLā€™)
  • 75ā„¦ coaxial cable terminated in a BNC connector at the dCS end

Another point to note is that dCS master clocks improve the performance of the system by providing a better optimized environment for the clock crystals, which are actually the same (exactly the same accuracy) as those in the DAC. This helps reduce jitter. This optimization includes everything from the physical environment (e.g. temperature, vibration) to the electrical environment (e.g. power supply, EMI).

The clocks in the DAC are already optimised to a very high degree and it takes quite a lot to improve on their jitter performance. So please understand that an external Master Clock doesnā€™t improve the performance just because it is external, but because of the freedom that allows in the realisation of the environment. An external Master Clock could well exhibit poorer jitter performance than the clock in the DAC if these aspects have not been addressed carefully.

1 Like
5

Andrew, I believe the Auralic clock frequencies are exact multiples of 44.1 and 48 kHz, but at the same time they are enormously higher than the frequencies of your own clocks. Iā€™m no electronic engineer but I assumed that would rule them out as incompatible (leaving aside the other points I mentioned). Is that the case?

6

So essentially the question is ā€is there an upper bound to the integer multiple?ā€ Good question, Presumably there is and it is programmed into the PLL in the FPGA. I donā€™t know. @James?

7

Yes, exactly. Iā€™m just curious, as I already have a dCS clock. They do seem to be more common though - I mean other manufacturers are making them - and perhaps dCS customers will want to try them. Itā€™s good to know what will be compatible and what wonā€™t. Of course ā€˜compatibleā€™ doesnā€™t mean an improvement in sound, as you pointed out.

8

Hi @Buizesmurf,

I took a look at the Auralic website. Regardless of whether the frequencies it outputs are compatible (our own crystals run at fs x 2^9, theirs at fs x 2^11) I see two potential and one clear showstopper here:

  • The specs donā€™t state if the Leo generates TTL or sine wave
  • The specs donā€™t state the impedance of the cable, and the connector supplied is incompatible (although the latter could be easily fixed)
  • The showstopper however is, as Simon noted above, the Leo requires the proprietary Lightning Link for the DAC to signal which frequency it requires. This turns the LEO into a completely proprietary system suitable only for use with the Auralic VEGA G2.1.
9

Yes, that is correct. The Auralic is not compatible with dCS. The maximum clock input rate on a dCS DAC is 176.4kHz / 192kHz, so rates higher than that (even if integer multiples) are not compatible. 192kHz is the upper limit.

Crystal rates are one thing, but if youā€™re sending a synchronisation signal between boxes (like a clock signal) higher signal frequencies does not mean better. Typically different signal frequencies generate different jitter spectrums, and for us, weā€™re very happy with the performance we achieve with 44.1kHz - 192kHz word clock signals. Itā€™s not a case of being able to encode more accurate timing information with more pulses - minimising jitter is key, and faster does not mean lower jitter.

It can be tricky to explain as you canā€™t really boil clocking performance down to a single number as there are far too many factors at play, but for us with our oscillators, our PLLs, our interconnects, our FPGA firmware, we see no benefit of going with higher rates.

And as Andrew says, there are some fundamental aspects of the Auralic that are a no-go for use with dCS, such as their Lightning Link.

3 Likes