Innuos - New Re-Clocker

hello @thyname
A 4TB Statement costs an eye watering ~$14,400 USD. A top Zenith 4TB + a Phoenix = $8,800 USD;

So, for approximately $5,600 USD more you get 2 chassis, a much more sporting power supply, heavy case work etc in the Statement. I assume they even tricked out further the implementation of the Re-clocker circuitry inside the Statement chassis;

If the Innuos deleted the Phoenix USB Re-clocker toroid, case, and all the extra parts to make a standalone unit, just focusing on the triple linear supply and USB Re-clocker circuit board itself (with the oven controlled oscillator on it) from the BOM of the Phoenix they would be in the hundreds of bucks of parts only;


The trick for them is integrating it in the Innuos Zen series chassis along with the same motherboard, which is a COTS PCBA from a PC vendor; The USB and Ethernet receptacles are directly mounted on the PCBA;

I have had custom Com Express Type 6 carrier boards designed for me such that the manufacturer removed exact USB and RJ-45 jacks (seen in boards like the InnuOs zen pc motherboard) in favor of parts from JST or SAMTEC, from which we would plug a mating board to wire connector and route the high speed differential signals over to secondary break out boards for other purposes;

InnuOs could totally do that here (add a larger overall VA transformer to power the USB stuff) and up the cost of the Zen series maybe $1,000 - $1,500 and still profit; Single chassis, lower costs for all; I can’t see them asking their PC supplier to add in the custom re-clocking circuitry but it is pretty trivial to break those signals out to a sub-system.

I am totally not picking on InnuOs at all; Just dreaming of ways to get the highest performance USB interface out there baked in a server that is around the ~5K mark;  

IMHO i’m sure the statement sounds incredible but boy is it turbo expensive!

I have a question about the re-clocker; So pretty much all modern DACs are async, where the DAC clocks data at its own rate from the USB host (as opposed to receiving bursts of data at whatever rate the HOST sends them);

When the reclocker was in the loop, async DACs still appeared as async to either the Aurender, windows PC, or InnuOs zen (as if the re-clocker was totally not there at all); I assume it's passing back the USB enumeration credentials back to the main USB HOST; 

So I guess what I have a question about is the role of the expensive oven controlled 3 parts per billion accuracy oscillator in the phoenix; I assume this clock pulls data from the HOST (music server), but the attached DAC is still operating in asycn mode pulling data according to its own clock-- unless the re-clocker re-configures the operating mode of the USB dac such that the re-clocker pushes data. 

Can someone please pipe in that has a clear understanding of what is going on here? TIA!

Some of it has to do with the DAC but in asynchronous mode it’s more of a push- pull. I am only a bit familiar with the DAC I have the Benchmark DAC3B it will follow sample rate changes. So in this case it would follow the Phoenix but only until it reaches the buffer, once there the DAC3B master clock takes it to send to the conversion subsystem it is the second asynchronous transfer that eliminates jitter in the transfer from USB to the conversion subsystem. In my case the Phoenix is just a worthless extra component reclocking data from the streamer to simply be reclocked by the Benchmark, other DACs might do it different. 

whelp the internet sure is a dandy thing

http://www.the-ear.net/how-to/power-supply-design-innuos-statement

looks like that’s exactly what Innuos did; They pulled the USB signal off the PC motherboard and piped it over to the reclocker;

~14K is a lot of scratch for power supplies, a small form factor computer, ssd's, teac optical drive, and spiffy cases; SHEESH!