Gustard U16 - the first USB Interface featuring ESS USB chip
Oct 22, 2019 at 9:27 PM Post #1,385 of 1,697
I think 1.77 is YMMV, Gustard actually said while 1.77 has addressed some bugs, due to sound quality issue, 1.77 will not be an official release.

That's true, YMMV..
So far no clicks and pops and everything seems alright so I'm staying with it.

PS: the 1.77 has reversed left and right channels on two different occasions, back to main firmware.
 
Last edited:
Oct 22, 2019 at 10:53 PM Post #1,386 of 1,697
Does anyone know how the U16 compares to an upgrade/modded SU-1 or stock SU-6?
 
Oct 23, 2019 at 7:33 PM Post #1,389 of 1,697
I am deciding between the two.

When attaching a master clock option to the Guastard U16, does it take that new sine wave signal for all of it’s clocking operations or does it still rely on the Accusilicon?

It seems like both the U16 and SU-6 use FPGA and PLL for clocking purposes. But one has a master clock option while the other has a temperature controlled monitoring system. It’s a tough choice.
 
Last edited:
Oct 23, 2019 at 9:58 PM Post #1,391 of 1,697
Actually I read it a little different. Correct me if I'm wrong.

"To ensure the highest quality digital audio signals and the best clock jitter performance, the U16 uses up to three FPGAs to implement USB audio interface clock management and the audio clock frequency synthesizers in the external 10M clock circuit."

Gustard uses three CME-HR FPGA's in total.
- (1) to synthesize a 24MHz clock for the USB input
- (2) to synthesize 44.1KHz and 48KHz base frequencies

"In the core circuit of the USB audio interface, an FPGA is used to implement the USB audio interface IC running in the Slave mode.That is, the USB audio interface IC receives the audio clock signal generated by the clock management FPGA and outputs only the digital audio data signal, thereby minimizing the digital signal interference introduced during the USB transmission process"

- USB audio interface receives audio clock signal from FPGA and outputs only the digital audio data signal thus minimizing it's own self clocking step

"The U16 has an external 10M clock input*, and the user can choose to access an external high-quality 10M clock reference source to further improve the sound quality.
* For security reasons, U16 does not automatically switch the clock source after accessing the external 10M clock. You must manually select the internal and external clock sources.
The audio clock synthesis circuit in the external 10M clock circuit uses two FPGAs corresponding to the audio main clock frequency synthesizer of the 44.1KHz and 48KHz base frequencies.
This design, on the one hand, avoids the crosstalk of the internal wiring of the FPGA; on the other hand, it also uses the relevant resources in the FPGA chip as efficiently as possible.
The frequency accuracy and jitter performance of the final synthesized audio main clock signal of this circuit have reached the limit performance level of the Current design"


Gustard U16 accepts external 10MHz reference clock signal.
- Internal and External clock operations must be manually selected
- Normally the two CME-HR FPGA's would take the ACCUSILICON 49.1520M and 45.1584M as a reference point to synthesize 44.1KHz and 48KHz base frequencies, but once an external clock is connected and manually selected, all three CME-HR FPGA's will use the 10MHz reference signal as the new reference point to synthesize the 24MHz, 44.1KHz, and 48KHz clock frequencies.

In short, I am pretty sure an external clock bypasses the internal clocks. With that said, this shows the technological advancements of Gustard U16 over Singxer SU-6 when it comes to offering an external clock option and properly implementing it. The only draw back is having to manually set it.

Aside from that, both the Gustard U16 and Singxer SU-6 have three things in common:
1. Reliant on femtosecond clock (ACCUSILICON & CRYSTEK) for clock synthesizing using PLL
2. FPGA implementation removes unnecessary isolation chip that was causing additional jitter at the I²S output.
3. Built-In FIFO is used for audio processing of buffered data before the output for higher bit-rate support

Differences:
1. External clock option
2. Power Supplies (Regular vs Super-capacitor)
3. Temperature Controlled Clock Monitoring System

Consensus: Although I am generalizing the details, the two have similar functions yet use different circuitry to achieve each their own. From the looks of it, the USB input on the SU-6 looks a little more advanced in terms of noise isolation but if one were to use an ISOREGEN or tX-USBultra before the unit, it may not matter. The power rails are clearly a difference with the SU-6 being a little more ahead. As for clock and synthesizer implementations, CRYSTEK is only a few femtosecond less than the ACCUSILICON. How much jitter these synthesizer chips produce between the two brands should be more important than the oscillators themselves. All in all, in order to know which sounds better we'll have to take their overall build into consideration. I am going to get myself both units and compare.

All I know is that the Gustard U16 and C16 together looks like a good deal. It would be an even better deal if they had a PLL audio clock synthesizer module that they sell that could be used in other clock areas of the digital chain that could connect to the C16. Sadly there isn't one, so one would have to rely on the more expensive SOtM sCLK-EX.
 
Oct 24, 2019 at 11:39 AM Post #1,393 of 1,697
Actually I read it a little different. Correct me if I'm wrong.

"To ensure the highest quality digital audio signals and the best clock jitter performance, the U16 uses up to three FPGAs to implement USB audio interface clock management and the audio clock frequency synthesizers in the external 10M clock circuit."

Gustard uses three CME-HR FPGA's in total.
- (1) to synthesize a 24MHz clock for the USB input
- (2) to synthesize 44.1KHz and 48KHz base frequencies

"In the core circuit of the USB audio interface, an FPGA is used to implement the USB audio interface IC running in the Slave mode.That is, the USB audio interface IC receives the audio clock signal generated by the clock management FPGA and outputs only the digital audio data signal, thereby minimizing the digital signal interference introduced during the USB transmission process"

- USB audio interface receives audio clock signal from FPGA and outputs only the digital audio data signal thus minimizing it's own self clocking step

"The U16 has an external 10M clock input*, and the user can choose to access an external high-quality 10M clock reference source to further improve the sound quality.
* For security reasons, U16 does not automatically switch the clock source after accessing the external 10M clock. You must manually select the internal and external clock sources.
The audio clock synthesis circuit in the external 10M clock circuit uses two FPGAs corresponding to the audio main clock frequency synthesizer of the 44.1KHz and 48KHz base frequencies.
This design, on the one hand, avoids the crosstalk of the internal wiring of the FPGA; on the other hand, it also uses the relevant resources in the FPGA chip as efficiently as possible.
The frequency accuracy and jitter performance of the final synthesized audio main clock signal of this circuit have reached the limit performance level of the Current design"


Gustard U16 accepts external 10MHz reference clock signal.
- Internal and External clock operations must be manually selected
- Normally the two CME-HR FPGA's would take the ACCUSILICON 49.1520M and 45.1584M as a reference point to synthesize 44.1KHz and 48KHz base frequencies, but once an external clock is connected and manually selected, all three CME-HR FPGA's will use the 10MHz reference signal as the new reference point to synthesize the 24MHz, 44.1KHz, and 48KHz clock frequencies.

In short, I am pretty sure an external clock bypasses the internal clocks. With that said, this shows the technological advancements of Gustard U16 over Singxer SU-6 when it comes to offering an external clock option and properly implementing it. The only draw back is having to manually set it.

Aside from that, both the Gustard U16 and Singxer SU-6 have three things in common:
1. Reliant on femtosecond clock (ACCUSILICON & CRYSTEK) for clock synthesizing using PLL
2. FPGA implementation removes unnecessary isolation chip that was causing additional jitter at the I²S output.
3. Built-In FIFO is used for audio processing of buffered data before the output for higher bit-rate support

Differences:
1. External clock option
2. Power Supplies (Regular vs Super-capacitor)
3. Temperature Controlled Clock Monitoring System

Consensus: Although I am generalizing the details, the two have similar functions yet use different circuitry to achieve each their own. From the looks of it, the USB input on the SU-6 looks a little more advanced in terms of noise isolation but if one were to use an ISOREGEN or tX-USBultra before the unit, it may not matter. The power rails are clearly a difference with the SU-6 being a little more ahead. As for clock and synthesizer implementations, CRYSTEK is only a few femtosecond less than the ACCUSILICON. How much jitter these synthesizer chips produce between the two brands should be more important than the oscillators themselves. All in all, in order to know which sounds better we'll have to take their overall build into consideration. I am going to get myself both units and compare.

All I know is that the Gustard U16 and C16 together looks like a good deal. It would be an even better deal if they had a PLL audio clock synthesizer module that they sell that could be used in other clock areas of the digital chain that could connect to the C16. Sadly there isn't one, so one would have to rely on the more expensive SOtM sCLK-EX.
Actually I read it a little different. Correct me if I'm wrong.

"To ensure the highest quality digital audio signals and the best clock jitter performance, the U16 uses up to three FPGAs to implement USB audio interface clock management and the audio clock frequency synthesizers in the external 10M clock circuit."

Gustard uses three CME-HR FPGA's in total.
- (1) to synthesize a 24MHz clock for the USB input
- (2) to synthesize 44.1KHz and 48KHz base frequencies

"In the core circuit of the USB audio interface, an FPGA is used to implement the USB audio interface IC running in the Slave mode.That is, the USB audio interface IC receives the audio clock signal generated by the clock management FPGA and outputs only the digital audio data signal, thereby minimizing the digital signal interference introduced during the USB transmission process"

- USB audio interface receives audio clock signal from FPGA and outputs only the digital audio data signal thus minimizing it's own self clocking step

"The U16 has an external 10M clock input*, and the user can choose to access an external high-quality 10M clock reference source to further improve the sound quality.
* For security reasons, U16 does not automatically switch the clock source after accessing the external 10M clock. You must manually select the internal and external clock sources.
The audio clock synthesis circuit in the external 10M clock circuit uses two FPGAs corresponding to the audio main clock frequency synthesizer of the 44.1KHz and 48KHz base frequencies.
This design, on the one hand, avoids the crosstalk of the internal wiring of the FPGA; on the other hand, it also uses the relevant resources in the FPGA chip as efficiently as possible.
The frequency accuracy and jitter performance of the final synthesized audio main clock signal of this circuit have reached the limit performance level of the Current design"


Gustard U16 accepts external 10MHz reference clock signal.
- Internal and External clock operations must be manually selected
- Normally the two CME-HR FPGA's would take the ACCUSILICON 49.1520M and 45.1584M as a reference point to synthesize 44.1KHz and 48KHz base frequencies, but once an external clock is connected and manually selected, all three CME-HR FPGA's will use the 10MHz reference signal as the new reference point to synthesize the 24MHz, 44.1KHz, and 48KHz clock frequencies.

In short, I am pretty sure an external clock bypasses the internal clocks. With that said, this shows the technological advancements of Gustard U16 over Singxer SU-6 when it comes to offering an external clock option and properly implementing it. The only draw back is having to manually set it.

Aside from that, both the Gustard U16 and Singxer SU-6 have three things in common:
1. Reliant on femtosecond clock (ACCUSILICON & CRYSTEK) for clock synthesizing using PLL
2. FPGA implementation removes unnecessary isolation chip that was causing additional jitter at the I²S output.
3. Built-In FIFO is used for audio processing of buffered data before the output for higher bit-rate support

Differences:
1. External clock option
2. Power Supplies (Regular vs Super-capacitor)
3. Temperature Controlled Clock Monitoring System

Consensus: Although I am generalizing the details, the two have similar functions yet use different circuitry to achieve each their own. From the looks of it, the USB input on the SU-6 looks a little more advanced in terms of noise isolation but if one were to use an ISOREGEN or tX-USBultra before the unit, it may not matter. The power rails are clearly a difference with the SU-6 being a little more ahead. As for clock and synthesizer implementations, CRYSTEK is only a few femtosecond less than the ACCUSILICON. How much jitter these synthesizer chips produce between the two brands should be more important than the oscillators themselves. All in all, in order to know which sounds better we'll have to take their overall build into consideration. I am going to get myself both units and compare.

All I know is that the Gustard U16 and C16 together looks like a good deal. It would be an even better deal if they had a PLL audio clock synthesizer module that they sell that could be used in other clock areas of the digital chain that could connect to the C16. Sadly there isn't one, so one would have to rely on the more expensive SOtM sCLK-EX.
 
Oct 24, 2019 at 11:52 AM Post #1,394 of 1,697
I have a combination of WD ext. HDD - Bryston BDP-1 - SOtM sCLK-EX - Gustard U16 (FW 1.4)/Gustard C16 - XLR - Auralic Vega - Auralic Taurus II - Sennheiser HD800/Focal Clear/Audeze LCD-X/Audeze LCD 3. I think I can hear the Vega for the first time with its full potential. The sound with all of the headphones is excellent. The contribution of the C16 is not negligible, on the contrary.
 
Oct 25, 2019 at 6:31 AM Post #1,395 of 1,697
Thanks for the share @zeid2

Do you feel like the Gustard C16 made a significant difference to your system?

Are you currently connecting it to your Gustard U16 and sCLK-EX board?


SC-Cut OCXO’s are better than femtosecond SPXO’s for sure however I didn’t know they were much better than the synthesizers that are on the sCLK-EX. The last time I checked their phase noise measurements they were pretty good. I believe even better than Crystek and Accusilicon.

One thing you should note is that the Gustard C16 outputs a 10MHz square wave. One of SOtM’s preferences is that they like using pure analog sine waves. Having said that, the sCLK-EX might perform better if it was fed that waveform. Certain chips might be optimized to receive a sine wave. Just food for thought.

The sCLK-EX’s synthesizers already performs pretty well. Having a Gustard C16 should help a little more with the phase noise but as previously mentioned, if the cable is too long or not good quality, I wouldn’t be surprised if they both sound the same due to signal loss increasing phase noise. From my experience I have found that Pasternack RG400 is the best coaxial cable for outside clock runs (SMB Male to SMA Female) and RG178 is the best for inside runs to the clock (SMA Male to u.FL [or MCX])
 
Last edited:

Users who are viewing this thread

Back
Top