Hi Mark,
That´s beyond my knowledge. The Singxer is very poor on documentation. But one reviewer said...
The crappy switching power supply of the Raspberry PI seemed to make no difference vs my linear supply when the Singxer SU-1 was used in the chain..
First, the linear power supply with a shielded Talema transformer and multi-stage LDO voltage regulation. This PSU sits on its own PCB separate from the rest of the device.
Singxer's board is a 4-layer PCB. A pair of Crystek CCHD-575 "femto" clocks handle timing, while two ISO7641 digital isolators provide galvanic isolation to keep noise separated. A custom Xilinx FPGA works its magic doing who-knows-what to help the signal stay clean and uncompromised.
That´s beyond my knowledge. The Singxer is very poor on documentation. But one reviewer said...
The crappy switching power supply of the Raspberry PI seemed to make no difference vs my linear supply when the Singxer SU-1 was used in the chain..
First, the linear power supply with a shielded Talema transformer and multi-stage LDO voltage regulation. This PSU sits on its own PCB separate from the rest of the device.
Singxer's board is a 4-layer PCB. A pair of Crystek CCHD-575 "femto" clocks handle timing, while two ISO7641 digital isolators provide galvanic isolation to keep noise separated. A custom Xilinx FPGA works its magic doing who-knows-what to help the signal stay clean and uncompromised.
Attachments
It's testable. The debate goes on because some people are unwilling to partake on or unaware of objective listening test.
From the Singxer web site. Probably translation with Google. I understand zero of this.
2.Used the high-speed whole isolation design, can be isolation between PC and interface of noise interference cleanly. SU-1 adopted a preeminent circuit design, using the high- speed CPLD/FPGA digital audio signal for separate after restructuring, so as to make the signal after reforming completely unaffected by isolation chip itself shaking.
3.SU-1 internal main board adopts the technology of the uniform source synchronous clock, CPLD/FPGA internal using special algorithm of clock signal and digital audio signal phase control strictly, can ensure that the digital audio signal after isolation chip will not appear extra distortion because of the time delay and phase difference.
2.Used the high-speed whole isolation design, can be isolation between PC and interface of noise interference cleanly. SU-1 adopted a preeminent circuit design, using the high- speed CPLD/FPGA digital audio signal for separate after restructuring, so as to make the signal after reforming completely unaffected by isolation chip itself shaking.
3.SU-1 internal main board adopts the technology of the uniform source synchronous clock, CPLD/FPGA internal using special algorithm of clock signal and digital audio signal phase control strictly, can ensure that the digital audio signal after isolation chip will not appear extra distortion because of the time delay and phase difference.
Okay. The USB/SPDIF board has its own power supply. The USB power is probably only used to power dirty side of the signal isolator chips, which is okay. The board probably operates the USB connection in asynchronous mode which allows it to use its own clocks (which appear to be pretty good quality) to clock out the SPDIF. Regarding the clocks, they will be somewhat sensitive to the quality of the power supply that feeds them. Best to have a dedicated high performance LDO for that located right next to the clocks. Can't see that part from the picture, but looks like they may only use multiple decoupling caps which would not be as good. However, we are talking about SPDIF here. There is probably a limit to how good its going to work just given the nature of SPDIF. You should use a good quality, short length, 75-ohm coax cable to connect the SPDIF output to your dac. Cable quality can affect the amount of jitter arriving at the dac since a poor cable will add more.
May I ask what type of dac you are using that requires you to use SPDIF?
May I ask what type of dac you are using that requires you to use SPDIF?
I have Chinese LITE DAC 83 with 4x BB PCM1704 balanced. I use VDH First Ultimate for SPDIF and 110ohm AES/EBU DIY VIABLUE NF-S2 Digital Cable. When I switch between them I hear no difference in SQ.
I have I2S on both, but don´t know how to implement, since there is no I2S specification for LITE DAC, and told by the expert that it was not worth it.
Is there any other way to connect my DAC?
I have I2S on both, but don´t know how to implement, since there is no I2S specification for LITE DAC, and told by the expert that it was not worth it.
Is there any other way to connect my DAC?
From Darko's review of DAC-83 sounds like it uses ASRC and upsampling to 24/192. That in theory should do a pretty good job of attenuating incoming jitter. Of course, it might not work quite as well in practice. Depends.
Galvanic isolation is the other thing your SPDIF box has that can help in many situations where system grounding is not well implemented. (Best home solution for that is probably a used 'Monster HTPS 7000 MkII' from ebay. Plug everything into separate ports in the back of it as they are labeled.)
Most external I2S inputs on dacs are intended for interfacing to HDMI, or to HDMI to I2S converters (although pinouts are not standardized that I know of). HDMI is not particularly low jitter either, it is mainly optimized for video.
My guess would be that going into the dac with I2S from a USB to I2S board like I2SoverUSB from JL Sounds (a very good one, and not expensive). However, figuring out how to make it work would probably take someone with more electronics background than you have.
Seems doubtful that Shanti will help much or possibly at all with your SPDIF box. Can't say for sure though. One would have to examine the internals more carefully to get a better idea.
Don't know what else to tell you. Sorry.
Galvanic isolation is the other thing your SPDIF box has that can help in many situations where system grounding is not well implemented. (Best home solution for that is probably a used 'Monster HTPS 7000 MkII' from ebay. Plug everything into separate ports in the back of it as they are labeled.)
Most external I2S inputs on dacs are intended for interfacing to HDMI, or to HDMI to I2S converters (although pinouts are not standardized that I know of). HDMI is not particularly low jitter either, it is mainly optimized for video.
My guess would be that going into the dac with I2S from a USB to I2S board like I2SoverUSB from JL Sounds (a very good one, and not expensive). However, figuring out how to make it work would probably take someone with more electronics background than you have.
Seems doubtful that Shanti will help much or possibly at all with your SPDIF box. Can't say for sure though. One would have to examine the internals more carefully to get a better idea.
Don't know what else to tell you. Sorry.
Thank you Mark for very thorough explanation. My upgrade itching, make me look for some easy and fast solutions.
Hard to admit but I will abandon the idea of LPSU at the USB side.
I might ending up with some kind of Intona solution after all. Just to exhaust the last available option.
Well, this is my disease of endless upgraditis.
Predrag
Hard to admit but I will abandon the idea of LPSU at the USB side.
I might ending up with some kind of Intona solution after all. Just to exhaust the last available option.
Well, this is my disease of endless upgraditis.
Predrag
A different, or an additional, galvanic isolator?
If you already have the USB to SPDIF board then you probably don't need another galvanic isolator. With the USB to SPDIF board, the problem, if there is one, would most likely be with its AC power input as a noise source. A USB isolator won't help with that. If the AC input is a problem, and if the power supply only outputs +5vdc then that might be an opportunity to use Shanti. However, you would have to do some modification of the power supply circuitry inside the USB to SPDIF board in order to use Shanti instead of the existing design.
Last edited:
Well, one of the reasons I have chosen Singxer is because it has descent PSU board. I don´t think it will be very beneficial to upgrade it with better PSU per se.
Now I understand that Singxer is good enough in rejecting noise and 5V LPSU will be redundant.
I ordered good 19V A LPSU for my Intel NUC, and hope this (at least) will improve SQ at least a notch or two, if not I am doomed again.
Now I understand that Singxer is good enough in rejecting noise and 5V LPSU will be redundant.
I ordered good 19V A LPSU for my Intel NUC, and hope this (at least) will improve SQ at least a notch or two, if not I am doomed again.
Hey HF4ME. I recommended Intona since that's what I use (USB 2.0 Industrial version), also with my Singxer SU-1. So I have same set-up as you. I wouldn't say it is necessary, but since it filters the USB to the Singxer + isolates the data signal that puts me at ease that the s/pdif is as good as it can be.
The Intona has voltage regulation, and the singxer has additional regulation again. Singxer only draws a couple 100 mA at most from USB, Shanti provides several ampere. So that's why I said Shanti is unnecessary.
Last edited:
I recommend reading this test over google translate Multitest externe 12-volt voedingen - Alpha-Audio
Sbooster is quite good. 'Farad Super' even better. Both are really expensive though. Shanti should be on Farad level, while supplying two separate 5V outputs instead of one, and being cheaper. It's basically your best bet for 5V power at great price.
iFi is cheaper and shows on testing. I would not buy iFi power supply.. Keces is also not stellar for the asking price, not bad but just not as competitive..
It's nice to see power supplies tested like this or we would never know which are good and which are not.
Hi @cdsgames and LPs specialists,
In the next few months I will have a USB sig + Rev dac rig.
I have one Salas L-adapter 5v 3a and 2 ultrabibs 5v 1,5a lps.
I am planing to power the usb sig with one L-adapter+ ubib and the rev dac with ubib.
I have also some Lt1083 lps I bought from audiophonics I can use too. They are a little bit stronger than ubib, but with lower quality...
I am just worried if the ubib would not be powerful enough for the rev dac.
I know that L-adapter is strong and very good quality, but Ubib at 1,5a....
Any help is welcome!
Thanks
Antônio
In the next few months I will have a USB sig + Rev dac rig.
I have one Salas L-adapter 5v 3a and 2 ultrabibs 5v 1,5a lps.
I am planing to power the usb sig with one L-adapter+ ubib and the rev dac with ubib.
I have also some Lt1083 lps I bought from audiophonics I can use too. They are a little bit stronger than ubib, but with lower quality...
I am just worried if the ubib would not be powerful enough for the rev dac.
I know that L-adapter is strong and very good quality, but Ubib at 1,5a....
Any help is welcome!
Thanks
Antônio
Last edited:
Again, IF the s/pdif clock is used by the D/A directly. If not it has an advantage in it's superior galvanic isolation (bar the typical 192 transfer limit).
//
+1 would like to know this too. But considering their katana I'm sure it will use synchronous master mode.
Thanks baten, I appreciate your comment and advise. Yes, the limitation of the USB by the specification is around 500mA max I think.
I will probably try Intona in the next couple of weeks.
Bravo!
I connected 5V power bank to the USB power leads and compared the SQ with non-splitted version (USB data+power coming from the NUC).
And guess what, the difference in SQ was ALWAYS in favour of splitted power.
The difference is not night and day but it is there. Less glare in highs, more musical.
I tried several times on and off and result was always the same. Being uncertain of the placebo effect, I lent a pair of fresh ears from my 17 y.o. son. He too, instantly noticed the difference in sound, on blind A/B test.
Why, if my Singxer SU-1, do the galvanic isolation? Well, my guess is that no galvanic isolation is 100% imune of the noise created in the USB cable travelling along the data leads. Maybe if I have an incredible good USB cable the situation would be different.
And finally, the 5V power bank for my test is far from perfect, as we know there is no 5V genuine chemical battery. What good clean 5V can do at this stage is still unanswered question.
In the next few days I will receive the 19V LPSU for my NUC from Audiophonics. I will present the results and try to stay objective.
128fs mode can be achieved by using a clk divider inside the es 9038
Rev has 2 settings ... 128fs (synch) and async
Async has dpll at 1 (44-192khs) and only 352 and 382Khz are synch ( meaning no clk division and still 128fs)
128fs mode has CLK division and only 352 and 382Khz in synch with no clock division