Article: Review | Innuos PhoenixUSB Reclocker

[Superdad]

5 hours ago, audiofool1982 said:

EX---> curious cable usb ---> ISO regen with LPS ---> Phasure Lush^2 ---> MSB analog Dac. 

Will the Phoenix USB reclocker over take the ISO regen in my setup?

 

It would, but realize that your ISO REGEN already has five separate Linear Tech LT3042 ultra-low noise regulators, a USB3.1 hub chip selected for best signal integrity, and an ultra-low-phase noise Crystek 575--plus true galvanic isolation via the costly Silanna ICE08USB chip. Combine that with the LPS you already use it with and the form factor allowing you to preserve SI and impedance match by plugging directly into the DAC with no additional cable, and I'd say you are already doing pretty well for about 1/10th the price. 

[I promise this will be my only comparative comment in this thread.]

[Superdad]

54 minutes ago, barrows said:

Respectfully, I would suggest though, that you might refrain from such definitive language as this:

 

54 minutes ago, barrows said:

The "quality" of USB output can be measured definitively through the eye pattern diagram combined with a spectral analysis of the USB power rail: these are the only relevant factors for USB transmission and subsequent audio performance (covering both the ultimate signal integrity and the noise factors).

 

Hi Barrows. Please take this in the playful Friday spirit I intend: You might do well to take your own advise regarding the above.

Because both an eye-pattern and VBUS spectral analysis will in fact not tell you all.  Proof?

Right there for you to check with two original Sonore microRendus, a v1.3 board and a v1.4 board, wherein the only difference between them is the latter's use of a low-phase-noise Crystek CCHD-575. The better clock is clearly heard, yes? But I challenge you to spot the difference in an eye-pattern test.

 

Sure, improved signal integrity common to both through the Rendu-series use of a USB hub chip at the output (and common to all USB "regenerators"--including the Phoenix--since JS started this trend back in 2014) can certainly be seen.  But the low-phase noise clock is not going to present in an eye-pattern; to understand its benefits one has to dig deeper (our 'white paper' points the direction).

 

Likewise, the 5VBUS lines (really the whole USB "ground" connection)  are a pathway for common-mode leakage currents which travel through all audio systems, so looking just at the noise level/spectrum on the +5VDC line will not expose those factors, which again can propagate ground-plane noise and increase receiver chip jitter by modulating the reference voltage used to determine clock thresholds (inside the DAC stuff, and I refer again to our paper).

 

Hope your Colorado weather is lovely this weekend and that you have something fun planned.

Cheers,

--Alex C.

[Superdad]

4 hours ago, austinpop said:

What I will say is that AES also benefits from reclocking and signal generation. I've experienced this with devices like the Mutec MC-3+, the SOtM dX-USB HD Ultra, and the Singxer SU-1, feeding several DACs.

 

Rajiv:

There is already enough misunderstanding about clocking out there so please be careful not to conflate audio-rate clocking (as done in DACs and DDCs--anything handling/generating S/PDIF, AES/EBU, I2S, or TOSLINK) with chip clocking for packet-data interfaces (Ethernet, USB, all computer chip clocks, etc.).

So completely different! The influence of sample-rate clocking (be it word-clock or bit-clock) is very direct.  

The influence of chip clocks (for USB, Ethernet, etc.) is vastly less direct and the mechanisms by which such can influence performance are completely different. (Reference our 'white paper' on that subject.)