Prima Lunacy? (SuperTubeClock)

I am planning on upgrading my tube amp when its current set of tubes start to go.  So I've been doing a bit of perusing online, and was really considering the Prima Luna integrated amps until I read about "the world's first tube-based data clocking device: the SuperTubeClock™"

Did I missing something or is this just about the most obvious snake oil sales job?  For what I am understand the Prima Luna Engineers are using a tube-based oscillator in lieu of a quartz crystal to generate the DAC chip clock signal.  Their blurb lists reduced jitter and noise as the advantages.  AFAIK the noise in a timing signal should be superfluous since it has two values 1 and 0, and anything in between (noise) is ignored.  If the clock signal "noise" is leaking into the final analog output, then there are big problems with the DAC chip.  With respect to jitter, I would expect the inaccuracies of a high-quality crystal oscillator to be measurable only in the nanosecond or picosecond range.  Can an analog-style oscillator really do better (and does it really matter)?

Another thing that stuck me is that the clock triode is soldered in, and there is mention of it lasting 5 - 10 years.  When it goes bad, do owners have to send their DACs to Prima Luna to be refitted with new - and possibly rare and otherwise unavailable - clock tubes?

Thoughts, anyone?
After reading the writeup on the "SuperTubeClock," and examining the waveform photos it provides after going to the URLs of the photos themselves (where they are presented in an easier to see form), I don’t consider this writeup to be BS or the SuperTubeClock to be snake oil. While the writeup raises some questions in my mind which are left unanswered/unexplained (see below), and although I certainly find a lot of other audio marketing literature to be BS from a technical standpoint and a turnoff when it comes to considering the product for purchase, I don’t in this case.

AFAIK the noise in a timing signal should be superfluous since it has two values 1 and 0, and anything in between (noise) is ignored. If the clock signal "noise" is leaking into the final analog output, then there are big problems with the DAC chip.

That is not correct. Noise on the clock signal applied to a DAC chip will result in short-term random or pseudo-random fluctuations in **when** transitions between 1 and 0 and/or 0 and 1 are sensed by the chip, causing the timing fluctuations that are referred to as jitter. (Generally speaking those transitions are sensed in the vicinity of their mid-point, i.e., roughly half-way between the two voltage states of the clock signal). And generally speaking that is widely recognized as being a significant issue in digital audio, to a greater or lesser extent depending on the particular design of course. It is not a matter of the noise "leaking through," it is a matter of the effects of the noise on timing.

In a thread a few months ago one of our technically astute members, @Kijanki, explained it this way:

Let me try to explain jitter. Imagine you play 1kHz sinewave recorded on your CD. Digital words of changing amplitude, representing sinewave, are converted in even intervals into analog values by D/A converter. You get analog 1kHz sinewave.

Now imagine that these time intervals are not exactly even, but are getting shorter and longer 50 times a second. Now you won’t get only 1kHz sinewave but also other frequencies, mainly 950Hz and 1050Hz called "sidebands". Distance from the main (root) frequency depends on the frequency of the interval change (jitter), while their amplitude is proportional to amount of interval change. These new sidebands have very small amplitude, but are not harmonically related to root frequency (1kHz) and that makes them still audible.

With many frequencies (music) there will be many sidebands - practically a noise added to music. Sidebands have small amplitude that is proportional to amplitude of the signal. This noise stops (is not detectable) when music stops playing. You can only hear it as lack of clarity in the music (since something was added).

Things the writeup leaves unexplained include the following:

1) **How** is the sine wave produced by the tube converted to a square wave? Presumably that is done by solid state devices, which may or may not introduce significant amounts of noise themselves and cause some amount of jitter regardless of how clean the signal provided by the tube may be. It would be nice if some indication of **overall** jitter performance were indicated.

2) Waveforms are shown for 8.4672 MHz and 42.2 MHz. What rate is actually being used in the D/A conversion? While the writeup and the photo for 42.2 MHz indicate that risetimes and falltimes of the square wave are about 1.67 ns, a corresponding figure is not presented for the 8.4672 MHz case. And it appears based on the waveform photo that those times are about 4 or 5 ns, considerably slower than at the higher speed.

In any event, I wouldn’t consider this writeup to be something that would dissuade me from considering this DAC, if I had a need for a new one. I certainly can’t say as much about a lot of other marketing literature I have seen from other manufacturers.

-- Al

Here’s two DACs similar in price. 
 the Mytek Brooklyn Bridge

Maximum output level / Impedance. 9.6Vrms / 73ohm
A-wtd S/N ratio (S/PDIF / USB) 112.9dB / 112.8dB
Distortion (1kHz, 0dBFs/–30dBFs) 0.0003% / 0.00017%
Distortion & Noise (20kHz, 0dBFs/–30dBFs) .0005% / 0.00035%
Freq. resp. (20Hz-20kHz/45kHz/90kHz) +0.0 to –0.0dB/–0.4dB/–1.9dB
Digital jitter (48kHz / 96kHz) <5psec / <10psec
Resolution (re. –100dBFs / –110dBFs) ±0.1dB / ±0.5dB
Power consumption15W (1W standby)

The Prima Luna EVO 100

Maximum output level / Impedance. 2.15Vrms / 2.35-12.2kohm
A-wtd S/N ratio (S/PDIF / USB) 104.9dB / 105.0dB
Distortion (1kHz, 0dBFs/–30dBFs) 1.07% / 0.025%
Distortion & Noise (20kHz, 0dBFs/–30dBFs) 1.03% / 0.026%
Freq. resp. (20Hz-20kHz/45kHz/90kHz) –1.6 to +0.1dB/–0.5dB/–7.4dB
Digital jitter (48kHz / 96kHz) 160psec / 80psec
Resolution (re. –100dBFs / –110dBFs) ±0.6dB / ±3.5dB
Power consumption 53W

Specs from Hi Fi News

It mentions in the review the Prima Luna uses a crystal as the Master clock the tube is used in he side bandwidth oscillator.
Impedance ... 2.35-12.2kohm

Ouch! That would certainly make it a poor match for the input impedances of a lot of solid state components, especially considering the wide variation of impedance as a function of frequency. Additional detail from the review:

Output impedance is excessively high at 2.38kohm through mid and treble, increasing further to 3.5kohm/100Hz and 12.1kohm/20Hz

A lot of the other differences seem expectable given that one DAC is tube-based and the other is solid state. Although the EVO’s jitter numbers don’t appear to be anything to write home about. Again from the review:

Jitter, in practice, is influenced by noise and other interference emanating from every corner of the digital circuit and is reduced from the 330psec measured with the ProLogue Eight to 160psec (48kHz) and 60psec (96kHz) in the EVO 100, with ±100Hz and ±200Hz PSU modulations the key components.

(The references to 100 Hz and 200 Hz would be 120 Hz and 240 Hz in the USA).

Thanks, @djones51. Regards,

-- Al

Post removed 
Wow, there are some seriously misguided views on some Chinese audio manufacturers. Very sad. If you did any research, you would find that the two brothers who started Line Magnetic have a great track record and that the LM facilities are far from your simplistic notion of asian sweatshops . Why are you so quick to judge people and their products, particularly in the hobby you partake, because of the country/politics they happen to be born into?