1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Singxer SU-1 Owners

Discussion in 'Dedicated Source Components' started by t bone, Mar 7, 2017.
First
 
Back
1 2
4 5 6 7 8 9 10 11 12 13
Next
 
Last
  1. oneguy
    My current chain is:
     
    MacBook Air (USB) > LH Labs 20 G (USB) > SU-1 (I2S) > Spring
     
    I don't have any experience with the Intona but I didn't like the Regen in my Pulse Infitiy chain.
     The 20G has it's own Li-ion 3100mAh battery which provides power for the output leg of the 20g box if needed. It also offers 10 Gbps up and down bandwidth and it's own signal regeneration and reclocking via femto clocks.
     
  2. Energy
    There's only 1 method that will work for my scenario.
     
    The iUSB will act as a floating power supply. The stock SU-1 power supply takes IEC, transformer, and linear voltage regulator to output a single 5V. The iUSB would just be in place of that, but with two 5V outputs supplying the USB input and mainboard. The USB cable from the computer will still plug into the Singer SU-1 as it normally would. (1 device method)
     
    The only performance boost here is the clean 0.1uV output noise from the iUSB 5V powering the SU-1's isolation section as well as the mainboard.
     
    Theoretically if run the data through the iUSB device I could further improve on galvanic isolation, however that cannot work as this device has an XMOS driver of it's own unlike the ISO Regen or Intona Isolator.. Two devices in series with its own independent drivers doesn't seem like it'll work.
     
     
    Or you can just drill in a hole to add another chassis mount DC output connector. As a cable manufacturer, I find two separate cables to be more rigid than one split cable.

    I would want to get myself an ISO Regen, but then that'll be additional cost for a LPS-1 on top of that.
    If I don't follow with my modification, I think I'm going to be cool with the stock power supply 3-5uV and just get an Intona to improve the 5V USB Input voltage noise. ah.... but again.. too many devices in series kills it for me.
     
    I still don't fully understand what signal regeneration is and why it's important.
     
  3. winders
    This is what your iUSB device is supposed to do when it regenerates USB:
     
    EyePattern.jpg
     
    The ISO Regen will do the same thing. Take a poor USB signal and regenerate so it is strong and clear like it should be. Anything you can do to get a strong and clean USB signal into the Singxer SU-1 is a good thing.
     
  4. inthere
    But doesn't the SU-1 do the same thing? I mean, take a poor USB signal and make it strong and clear?
     
    I think that's probably why some are get barely noticeable results with the SU-1+(something else)...........the SU-1 does such a great job that there's not much left to "clean"
     
  5. Energy
    Thats so informative. Thank you so much.

    I've actually come across the whole eye pattern thing before. Same for regenerate, impedance match, and rebalancing but never took the time to figure it out.

    So from what you're saying, if I was to put one of these iUSB circuit boards into my Singxer in place of the stock power suply, I shouldn't use it for power output only, but should have my USB data run through it for the benefits listed above.

    Hmmmmm.. this changes everything.

    In that case i'll have to contact iFi Audio to see if I can use both the iUSB and Singxer SU-1 at the same time even though they have independent drivers.

    I get it now why you're more keen to recommend me the ISO REGEN. Its because if I got the Intona Isolator I'd have isolation and clean power, but no regeneration.

    T Bone used the Singxer SU-1 + Intona Isolator and found a mild improvement not worth $250, but I'm at this point I'm convinced that using a iUSB Power Supply or ISO REGEN in place of the Intona Isolator will offer much more significant gains.

    This is great. You do your mod and I do mine. People can pick whichever route they wish if they ever want more from the SU-1.

    I'm only using the iFi because I can find a used one for $100-130 and fits inside of the Singxer SU-1 where the stock power supply would be. It also only outputs 0.1uV noise. Can't get much better than that and I haven't seen too many linear power supplies with that number. Single digit uV maybe, but not decimal.

    You also have to factor in the cost of the ISO REGEN alongside a LPS-1. Pricey pricey.

    The Singxer SU-1 is only $350-400. I wouldn't want a mod to be more than half its cost. Diminishing returns becomes a big thing. YMMV.
     
  6. winders

    No. It's job is to convert the USB signal into another signal (AES/EBU, SPDIF Coax, I2S). Computers, generally speaking, put out really bad USB signals. Then you have the power issues. That's why isolators tend to improve things.
     
  7. winders

    Which iUSB do you have? Do you have one of the newer units or the older unit. I am not sure what the older one does in regards to regeneration, etc.
     
  8. winders
    Of course, there are some people who say none of this matters and that you can't hear any difference. That you should just use a USB cable and plug that straight into your DAC. After all, it's all digital, right?
     
  9. Energy

    The Singxer SU-1 job is to isolate itself from dirty computer noise.

    First it receives the packets of USB data, makes sure every bit is there (1's and 0's) by running it through the XMOS chip, a process called asynchronous. It then takes the repackaged USB signal and isolates its ground and power from the computer, a process called galvanic isolation. This process however is powered by the computers noisy 5V USB output. The process can perform better with cleaner voltage. Once the signal is isolated, it moves onto the mainboard in which it use an entire new 5V and ground provided by the internal power supply. It is then reclocked by the femto clock and converted into various "flexible" outputs like i2S, AES, and SPDIF.

    As stated earlier, there are 3 problems.
    1. USB Input voltage from the computer is noisy and can be improved in order for the isolator on the Singxer to perform optimally.
    2. Power supply voltage to the mainboard, although good, can still be improved. This will help lower the noise floor for the reclocking side of the device.
    3. USB data signal entering the Singxer is not in their "best" condition. For an ideal USB signal, it needs to be impedance matched (90 ohm), signal level must be properly restored and rebalanced, and removing further common mode noise is a good extra.

    This is why placing a device in series before the Singer SU-1 such as an iFi iUSB or ISO REGEN will help prepare the "PERFECT" signal. In theory all this accounts to the decrease of jitter.


    I only have the first one that is USB 2.0. Last time I checked they were the same, just the newer one is rated for USB 3.0 and comes with other gadgets.

    I'll ask them again.

    If it doesn't do that I'll go for the iFi Nano iUSB 3.0.
     
  10. winders
     
    Not really, It's job is to convert the USB signal to another signal. In the process it tries to do other things. It tries to isolate the "dirty" side (USB input side) from the "clean" side (output side).  As you state, the "dirty" side is powered by USB bus power and the "clean" side is powered by a 5 VDC source. From what I understand, the isolation chips used aren't able to achieve true isolation.
     
    That's why some people use a true galvanic isolator between the computer and the SU-1. None of the iFi products listed on the web site is a galvanic isolator. That product appears to be in the works.
     
  11. Energy
     
    Ah thanks for catching my flat out statement sentence, but yeah, I didn't mean it like that. It's job is to convert to another signal (generally better) with better clocks, but it does isolate dirty noise in the process. It's 40% or so of what it does.
     
    And I heard the same thing. That explains why adding the Intona Isolator gives a subtle improvement as it's finally able to achieve full isolation.
     
    I did some research. Turns out only the iFi iUSB 3.0 offers the Rebalance,Reclock, and Regenerate. iUSB 2.0 does not. Both however offers IsoEarth which according to iFi Audio galvanically isolates 95% of the problem. Also the iUSB 3.0 Nano is out of the picture as it doesn't offer IsoEarth (galvanic isolation) at all.
     
    "Not all galvanic isolation is the same.  Some DACs have proper implementation, some not. In practice, the micro iUSB3.0 delivers 95% of what galvanic isolation would, without the extra cost, speed limits etc. et al. Working on an assumption that genuine galvanic isolation is 100%, iUSB3.0 with IsoGround is 95% at much less cost and full USB 3.0 Superspeed." (source: http://www.head-fi.org/t/779504/ifi-iusb3-0-launch-names-back-in-the-hat-and-new-2-winner-page-24/375#post_12005887)
     
    So the good thing is it does 95% of the issue and so maybe the Singxer SU-1 will clean up the rest. But the problem is, if the isolation chips like you say already doesn't manage that, then that last 5% seem like it'll remain there. 
     
    I'm now interested in the ISO REGEN if it can do a 100% true galvanic isolation without bottlenecks in other places.
     
  12. rocky500
    Which driver is better with the Singxer SU-1?
    I had a Holo Dac first with the latest Holo Driver. Then connected the Singxer which automatically setup using the Holo driver.
    I thought I read somewhere that the Holo driver goes higher in native DSD, so did not bother uninstalling and putting on the Singxer driver.
     
  13. T Bone
    A word of advice on the drivers.  Since both the Holo and the SU-1 are XMOS based devices they are functionally equivalent.  As such, you do NOT want to have both drivers installed simultaneously.  I did this once and got my PC very screwed up!  
    I spoke to Tim @ Kitsune and he recommends that you use Holo driver over the Singxer driver.  The Holo driver is "signed" and tested.  The SU-1 driver is essentially a generic XMOS driver.  

    I am fairly certain the driver has zero bearing on which DSD sample rates are supported.  ....that's the job of the devices firmware, not the driver installed on the PC.
     
    rocky500 likes this.
  14. Tboooe
    So does that mean Letner was willing to make a DC input one?  In the past when I have asked him he said he had no plans for offering a DC input version.  So I assume he is going to just add a DC barrel socket to the back?
     
  15. Tboooe
     
     
    The other main difference between the Holo and Singxer driver is the ability to play native DSD vs DoP.  That may not matter to people but for me I use HQP and wanted to only use native DSD.
     
    rocky500 likes this.
First
 
Back
1 2
4 5 6 7 8 9 10 11 12 13
Next
 
Last

Share This Page