[Chodi]

  I'm considering letting my Gustard U12 go, and have put up a interest check on the For Sale/Trade forum here:
 
http://www.head-fi.org/t/740647/ic-gustard-u12-usb-to-spdif-interface
 
If anyone is interested in getting their hands on this without having to deal with shipping from China, let me know!

The Gustard takes a few days for the clocks to settle down. You might wnat to leave it on an burn in a few days before jumping ship. It really changes a lot in a few days. You also might want to try a full version of that Thesycon driver. The one supplied is really basic.

 

[Alexsander]

Very interesting devicу this Gustard U12:
 
"Newest XMOS chip and uses 48MHz oscillator to asynchronous reclock usb audio data to I2S line, old XMOS uses only 13Mhz oscillator, over 4x better jitter rejection theoretically"
Display, PSU inside, AES/EBU, i2s , opt and coax and Built-in smart switch, plug in USB automatically!!!
 
But Digital Transformer Pulse PE-65612NL (limits 7 mbps 96kHz)!!!  it is necessary remove and istall Murata DA101C!!!
 
 
P.S.
Full Technical Parameters:
 
Laser marking, never fade
Built-in smart switch, plug in USB automatically
Two 0.1PPM precision crystal 45.1584M 49.1520M
International voltage AC115V-230V, the need to manually adjust
Rigorous design, reasonable layout, elegant lines, solid materials.
Figures show that a mesh understand. DSD support 128, PCM supports 384K
Aluminum chassis, the drawing process, the body all black, black and silver optional panel
XMOS - U8 chip, support 32BIT / 500MIPS digital signal processing unit, the inside has eight cores.
Interface from left to right are: power input AES / EBU output coaxial output IIS output optical output USB input
Machine size (not including protruding parts): Width 160mm × 50mm × depth 120mm high
Package Size: Width 225mm * 80mm * Deep 190mm high
Weight: Net Weight 1KG, including packaging 1.3KG
 
Features:
Increase the screen display, the sampling rate at a glance
Use XMOS programs to support 32B 384khz asynchronous 2.0 (USB recommended that ASIO output)
Using an independent power supply. Built-in sealing ring cattle, multiple sets of multi-level regulation. About the importance of independent power, I believe we all know.
Optimize the design of the audio clock circuit using 0.1PPM precision active crystal to provide a more perfect clock signal to digital audio. Provided for by the two crystals were 44.1 / 88.2 / 176.4 / 352.8Khz and 48/96/192 / 384Khz series rate signal only after the selected clock rate selected work, work another one audio clock stop, put an end to the mutual interference may bring work simultaneously when the two clocks.
A total of four kinds of output: Coaxial (RCA), optical fiber, AES / EBU, IIS (HDMI port)
Coaxial fiber AES / EBU output formats IEC60958 and AES3-r2009 standard limits the maximum support 24Bit / 192kHz; IIS can transmit 32B 384K
Smart switch, USB has a current interface to boot. Interface with the computer switch and switch machine (computer to be set off USB power failure), or unplug the USB cable directly to the shutdown.
 
Description
Gustard U12 digital interface XMOS program complete independent linear power interface 0.1PPM crystal support 32B 384K

U12 uses the company used in professional audio field XMOS latest chip U8, inside there are eight cores, with powerful multi-core multi-thread processing capability, is the industry's very good USB Audio Class 2.0 digital audio solutions.

 
 
 

 

[Alexsander]

Maybe it is worth replacing my current  setup?:
 
using modding Leaf Audio XMOS 384K/32bit PCM5102 DAC (Xmos version 8L6C5 GT1325L1) as usb converter USb-to-SPDIF with external power TeraDak U9VA Linear Low noise Power Supply + Power Supply A&B Cardas pure copper DC cable 1.5ft Comp: TinyMPD.XMOS 384K/32bit PCM5102 DAC
(My upgrade this device:
At first remove digital transformer Pulse PE-65612NL (limits 7 mbps 96kHz)->Murata DA101C, Second remove low quallity RCA jack -> Neutrik NF-2D-2, wire and silver solder is AudioQuest)
http://photo.qip.ru/users/audioquest/200709111/213720907/#mainImageLink
 
whether will give the best sound Gustard U12?

 

[rb2013]

Maybe it is worth replacing my current  setup?:

using modding Leaf Audio XMOS 384K/32bit PCM5102 DAC (Xmos version 8L6C5 GT1325L1) as usb converter USb-to-SPDIF with external power [COLOR=333333]TeraDak U9VA Linear Low noise Power Supply + Power Supply A&B Cardas pure copper DC cable 1.5ft Comp: TinyMPD.[/COLOR]XMOS 384K/32bit PCM5102 DAC
([COLOR=333333]My upgrade this device:[/COLOR]
[COLOR=333333]At first remove digital transformer Pulse PE-65612NL (limits 7 mbps 96kHz)->Murata DA101C, Second remove low quallity RCA jack -> Neutrik NF-2D-2, wire and silver solder is AudioQuest)[/COLOR]
http://photo.qip.ru/users/audioquest/200709111/213720907/#mainImageLink

whether will give the best sound Gustard U12?

It looks like the power supply and filtering is better in the Gustard U12. The more I'm reading alot about the issues of ps ripple on digital clocks, good ps regulation and filtering a particular power static caused by fast alternating power devices like motors (even small ones like cpu, pc ps, video cards use for cooling) and switching power supplies (the worst offends).

Edit typos

 

[stjj89]

  The Gustard takes a few days for the clocks to settle down. You might wnat to leave it on an burn in a few days before jumping ship. It really changes a lot in a few days. You also might want to try a full version of that Thesycon driver. The one supplied is really basic.

 
Thanks for the advice! I let it burn in for about 2 days by playing music using the ASIO driver through it and to my DAC, but haven't noticed big differences. I am very interesting in hearing how the Thesycon driver sounds like though. However, when I initially tried installing it, setup.exe would keep failing (and telling me to re-run the setup). The specific error is 
 
ERROR: Pre-installation of 'C:\Program Files\WaveIO\USBAudio_Driver\WaveIOks.inf' failed.
Please run setup again!
 
Luckit WaveIO Drivers (v2.23.0).exe would not recognize my device and keep telling me to plug it in (even though it already is). Any advice on how to get the installation to work?

 

[Wynnytsky]

Is Alexsander correct about there being a 96khz sample rate limit?
 
Has anyone tried Thesycon v1.6.1 on the U12?  If it works then stjj89 could use those instead of Chodi's drivers.
 
Jamiee - what did you think of the jump from U10 to U12?
 
I finally got a U12 and I had both the U10 and U12 plugged into seperate USB ports.  To A/B I would switch output devices in JRiver and move the AES cable between units.  The new 2.23 drivers (not Thesycon) see the U10 and U12, though ASIO was only avail for the U12.  The first minutes were pretty bad for the U12.  My broken in U10 sounds very relaxed in comparison to the U12.  Even though the new ASIO driver is not from Thesycon, it still sounds better than WASAPI, though I recall the ASIO/WASAPI performance gap being larger on the U10.
 
It's premature for me to report my findings because mine needs more break in time, but this is what I've heard so far:
U10 is all about scale -- bigger than life concert imaging like maggies or lascalas -- it's what set the Gustard apart from other products and why I liked it so much.  Scaled/magnified imaging (ex: Gustard U10) and focused/pinched imaging (ex: AudioPhilleo) are opposites of each other (I'm certain there is an optical lens analogy to be made here).  While either may be appealing to the ear, reality is somewhere in the middle.  I found Musiland and Amanero to be more honest in that respect.  The U12 has taken a significant step toward focus and I heard a texture in the upper vocals that my U10 hadn't revealed.  I really thought the U12 was only adding an LED display but there is definitely an audible change and I'm eager to hear where my U12 is going as it breaks in.

 

[genclaymore]

   
Thanks for the advice! I let it burn in for about 2 days by playing music using the ASIO driver through it and to my DAC, but haven't noticed big differences. I am very interesting in hearing how the Thesycon driver sounds like though. However, when I initially tried installing it, setup.exe would keep failing (and telling me to re-run the setup). The specific error is 
 
ERROR: Pre-installation of 'C:\Program Files\WaveIO\USBAudio_Driver\WaveIOks.inf' failed.
Please run setup again!
 
Luckit WaveIO Drivers (v2.23.0).exe would not recognize my device and keep telling me to plug it in (even though it already is). Any advice on how to get the installation to work?

You could do manually edit the ini with the info of your device to make it install. But you might have to do it with driver signature disable.

 

[sandab]

 
"Newest XMOS chip and uses 48MHz oscillator to asynchronous reclock usb audio data to I2S line, old XMOS uses only 13Mhz oscillator, over 4x better jitter rejection theoretically"
Display, PSU inside, AES/EBU, i2s , opt and coax and Built-in smart switch, plug in USB automatically!!!
 

This is only on the source clock though, and scaling down a higher frequency clock with a divider or rotating shift register doesn't necessarily produce a reduction in jitter.  In fact, this scaling can easily add jitter of its own, especially if using CMOS instead of a current-mode logic like ECL or such.  But in the grand scheme of things the source clock is hardly a problem; the real source of jitter is noise on the transmission line, reflections, nonlinear phase, etc, that results in jitter on the receive side.  Differential signaling (IIS, AES3) is clearly of potential benefit here over single-ended coax, but it's not immune.  How many ps of jitter there is is kind of unimportant.  (FemtoClock BTW is a trademark for a series of integrated clock generators and not really about real-world measured performance in a home or office environment.)

 

[rb2013]

This is only on the source clock though, and scaling down a higher frequency clock with a divider or rotating shift register doesn't necessarily produce a reduction in jitter.  In fact, this scaling can easily add jitter of its own, especially if using CMOS instead of a current-mode logic like ECL or such.  But in the grand scheme of things the source clock is hardly a problem; the real source of jitter is noise on the transmission line, reflections, nonlinear phase, etc, that results in jitter on the receive side.  Differential signaling (IIS, AES3) is clearly of potential benefit here over single-ended coax, but it's not immune.  How many ps of jitter there is is kind of unimportant.  (FemtoClock BTW is a trademark for a series of integrated clock generators and not really about real-world measured performance in a home or office environment.)

See my comments on the effect on clocking by power supply ripple and noise. These clocks may test well in electricly pure conditions, this isn't real world. There is a tremdeous amount of grunge in our nieghborhood power system. Like when the neighbor is using his power saw, you hear a whizzing sound on your radio. That's just one source of this contamination, digital devices are notorious for feeding into the grid this form of noise. Digital clocks depend on a very pure power source. Any power irregulatory or noise cause micro mistimings that produce 'jitter'. That's why that name was coined for one of digitals's distortions. But there are others - liked the mentioned phase shift issues, digital 'echoing' or reflections. The engineers at Audio Sensibility attest that the optimal length of a rca spdif cable is 1.5m, this length minimizes digital reflections. But there are a host of others - like dynamic impedence misalignments between cable and socket from the ideal 75ohms.

 

[rb2013]

Thanks for the advice! I let it burn in for about 2 days by playing music using the ASIO driver through it and to my DAC, but haven't noticed big differences. I am very interesting in hearing how the Thesycon driver sounds like though. However, when I initially tried installing it, setup.exe would keep failing (and telling me to re-run the setup). The specific error is 

ERROR: Pre-installation of 'C:\Program Files\WaveIO\USBAudio_Driver\WaveIOks.inf' failed.

Please run setup again

 
Luckit WaveIO Drivers (v2.23.0).exe would not recognize my device and keep telling me to plug it in (even though it already is). Any advice on how to get the installation to work?

Try an uninstall of any previous ASIO programs. Then reinstall the WaveIO. It can conflict with other ASIOs.

 

[rb2013]

Is Alexsander correct about there being a 96khz sample rate limit?

Has anyone tried Thesycon v1.6.1 on the U12?  If it works then stjj89 could use those instead of Chodi's drivers.

Jamiee - what did you think of the jump from U10 to U12?

I finally got a U12 and I had both the U10 and U12 plugged into seperate USB ports.  To A/B I would switch output devices in JRiver and move the AES cable between units.  The new 2.23 drivers (not Thesycon) see the U10 and U12, though ASIO was only avail for the U12.  The first minutes were pretty bad for the U12.  My broken in U10 sounds very relaxed in comparison to the U12.  Even though the new ASIO driver is not from Thesycon, it still sounds better than WASAPI, though I recall the ASIO/WASAPI performance gap being larger on the U10.

It's premature for me to report my findings because mine needs more break in time, but this is what I've heard so far:
U10 is all about scale -- bigger than life concert imaging like maggies or lascalas -- it's what set the Gustard apart from other products and why I liked it so much.  Scaled/magnified imaging (ex: Gustard U10) and focused/pinched imaging (ex: AudioPhilleo) are opposites of each other (I'm certain there is an optical lens analogy to be made here).  While either may be appealing to the ear, reality is somewhere in the middle.  I found Musiland and Amanero to be more honest in that respect.  The U12 has taken a significant step toward focus and I heard a texture in the upper vocals that my U10 hadn't revealed.  I really thought the U12 was only adding an LED display but there is definitely an audible change and I'm eager to hear where my U12 is going as it breaks in.

I've been running 192k for weeks without issue - that's 24/7. Not a single error message, after trillions on bits sent and received on the U12, not one hiccup, unlock, or freeze.

Thanks Chodi! That Tspy tool is awesome - tracks everthing in real time.

Swapping the blue Phillips caps for better Pani FCs helped to smooth things out a bit. Better ripple filtering.

 

[sandab]

Oh yeah, you have to have clean AC.  I assumed that was taken care of already.  People always go on about how noisy it is, but when I put a transformer on it here and scoped it to check the spectrum, it looked pretty clean to me.  With my cheap Furman PST8D it looks flawless - no special snake oil required, but taking a look with a scope with spectrum view is the only way to know.

 

[rb2013]

  Oh yeah, you have to have clean AC.  I assumed that was taken care of already.  People always go on about how noisy it is, but when I put a transformer on it here and scoped it to check the spectrum, it looked pretty clean to me.  With my cheap Furman PST8D it looks flawless - no special snake oil required, but taking a look with a scope with spectrum view is the only way to know.

I wouldn't assume to much.  Not here - even worse the voltage itself would fluctuate during the day.  I once had a Monster power conditioner on my HT system.  Would watch the voltage change constantly.  But that's not what I'm talking about here - it's the noise generated by your pc itself.  Fans, monitors, pc ps, also those switching ps wall warts we all have plugged in all over.  Not to mention the intermittent grunge poduced by the nieghbor's vacumm cleaner, power saws, etc...
 
Some interesting stuff:
 
A typical power supply of notebook PC
Sources of disturbances (ripple & noise) inside a computer:

1. Backlight of computer monitors: The necessary high voltage of approx. 900 V for LCD/TFT monitors is generated by inverters which produce ripple and noise.

2. Hard disks (internal + external): The actuators of the recording/reading heads produce interference voltage with each positioning or stop move.
3. Clock generators and chips like e.g. the CPU, northbridge, RAM or graphics adapter cause disturbances and noise.
4. Power supplies of notebooks and desktop PCs are not designed for lowest interference voltage and noise, since this is extra costs, not necessarily needed.

5. DC/DC converter producing up to 100mV ripple and noise.
and so on...

100mHz digital scope, one vertical square are 10mV - this shows 50 - 60mV
 
 
Interference voltage and noise of USB power from a computer. 
Even the use of a battery-powered notebook brings little help.
A typical picture of 5 Volt USB power with 40 mV ripple & noise from a desktop PC without running access of the hard disk.
 

100mHz digital scope, one vertical square is 10mV - this shows 40 - 50mV
 

A typical picture of 5 Volt USB power from a desktop PC with running access of the hard disk. 350mHz Analogscope, one vertical square is 10mV - this shows 60 - 90mV 

 

[rb2013]

n CD players this is a particularly troublesome issue - the servo motors used to align the laser head to the irregular CD pit 'groove'.  Those servo motors, unless very carefully designed feed this grunge back through the ps to the clocks and dac portion of the CD player.
 
While digitalized music recordings fixed a host of issues related to analog (like LP surface noise, tangential tracking issues, etc...)  They introduced a whole host of new issues, that we're finally getting around to addressing.  The early CD players were horrible - I had a few.  Rather then the marketed 'perfect sound', they were bleached, etchy, glaring, lacking any real life.  I way preferred my analog system.  But today well done digital can be really good.  Unfortunately, with the movement away from R2R dacs to the newer cheaper to produce designs we've taken a step backwards. Unless the audio engineers are very careful about their designs.  They have plenty of boom/sizzle - but many lack a lifelike tonal richness. 
 
Here is a good read on jitter:
 
http://www.tnt-audio.com/clinica/jitter1_e.html
 
For DACs it's the issues regarding delta-sigma and multi-bit dac designs vs the older R2R.
 
http://www.mother-of-tone.com/conversion.htm
 
http://funwithaudio.blogspot.com/2012/01/today-in-electronics-everything-is-made.html

 

[sandab]

 
Interference voltage and noise of USB power from a computer. 
Even the use of a battery-powered notebook brings little help.
A typical picture of 5 Volt USB power with 40 mV ripple & noise from a desktop PC without running access of the hard disk.
 

100mHz digital scope, one vertical square is 10mV - this shows 40 - 50mV

Interesting.  I just got one of these - http://www.mouser.com/Search/ProductDetail.aspx?R=USB-TERMvirtualkey58150000virtualkey992-USB-TERM - to scope the +5V out from my MacBook Pro with different cables of different length, to see what goes on and to see what exactly the Wyrd does and how well it cleans it up.  I'll be checking it open (floating) and with 220k, 2.2k, and 5V/500mA = 10ohm loads for each cable.  Both spectrally and with a DMM for voltage.  My hypothesis is that a "good" audio-grade USB cable has relatively high impedance in the +5V/gnd and the voltage will drop under load, making it unsuitable for powering, but fine for a high-Z reference input (like XMOS parts seem to want it for) - and, so, intrinsically choked.  In general, if the noise goes down with load without affecting the voltage too badly, shunting may be an option.  I may also add noise generation to see what impact it really has and what's audible.
 
This of course is completely different from AC noise.  Also, a computer could be extremely noisy internally yet have a clean USB out, if designed to that end.