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


Discussion in 'High-end Audio Forum' started by magiccabbage, May 14, 2015.
216 217 218 219 220 221 222 223 224 225
227 228 229 230 231 232 233 234 235 236
  1. romaz
    I used to own the PS Audio DirectStream and sold it shortly after the Yale firmware was released last year.  I tried all the various firmware that were available as this DAC allowed you to go back and forth so that you could decide which presentation you liked the best.  In that sense, this DAC was a chameleon and consequently had somewhat of identity issue as you could choose the older firmware that was considerably darker or upgrade to the latest firmware that was tipped more to the treble but was also better resolved. It was almost like tube rolling.
    Ted Smith, the "Rob Watts" equivalent for PS Audio, is a brilliant man and a nice guy as well.  Like Rob, he is very active and helpful to answer questions on the PS Audio forum and he is a large reason for the success of PS Audio.  Like Rob, he very much thinks outside of the box but he and Rob have philosophical differences regarding their FPGA DACs and it will be readily evident once you take a listen.  Ted upsamples everything to DSD and as we know, Rob's DACs do not do that.  My DirectStream had a soft/smooth signature but lacked in detail.  It also sounded very flat and when this didn't improve when Yale came out, I realized it was time to move on.
  2. bigfatpaulie
    I guess the reason I ask is because DAVE seems to be pretty widely renowned as the PCM DAC right now.  DSD is its achilles heel which is where the DS is at it's best.  But if DSD is game, a 'chipless' Lampizator is likely easily king.
  3. shuttlepod

    I don't disagree with bigfatpaulie. I would only add that DSD still sounds pretty damn good on the DAVE (better than it ever sounded on my Ayre QB-9 DSD, although to be fair Ayre is not known among DSD fans as a top DAC). High quality PCM recordings are truly amazing on DAVE, and PCM in general is pretty great. Unless DSD is the format you prefer and you have lots and lots of DSD recordings, I'd give DAVE a serious listen.
  4. Rob Watts
    OK I think perhaps you have misunderstood. There are two uses of capacitors (caps) - DC blocking caps, which prevent DC from flowing from one stage to another, and filter caps. DC blocking caps typically have a value range of 4.7 uF to 470 uF. These caps are electrolytic (small but suffer from lots of LF distortion, have large leakage currents and need time to form the dielectric hence break in) then you have tantalum which are about 10 times better than electrolytic's but more expensive. Then you have film and foil caps, and the best of these are PTFE (which are hideously expensive but lowest form of dielectric absorption and no distortion) then polypropylene (which have very low levels of dielectric absorption, and no distortion) then cheaper films that aren't as good.
    The problem with film and foil (apart from cost) is physical size, even for values as low as 4.7 uF and you simply can't get big values. The large size means much more noise pick-up. So with Dave I use NO coupling caps at all, everything is DC coupled. To remove the DC from the OP, I then use a servo. All conventional servos are analogue, and this means that the coupling cap is replaced with an integrator cap, which will still add distortion and noise when the trimming DC is added.
    So what I do is to take the output from the DC integrator, which has distortion from the integrator cap, and noise from the analogue integrator, then I use an ADC and feed this digital data back into the FPGA. The DC trimming data is then very heavily filtered digitally, so all of the noise and distortion is removed completely, so we end up with a very small very gently moving trimming data that is added to the noise shaper input, and when it gets to the analogue OP we end up with no DC. Now when Dave is first switched on, the DC servo acquires the data, and then writes to FLASH the actual value of the DC trim that is required so when you power Dave up again it starts up with the correct value. But even with this in place, it still takes 20 seconds for the system to remove the DC, and that's why there is a large turn on and off delay when you switch modes or turn on the unit.
    So the benefits of this rather complex system are that there is no DC coupling caps in the signal path at all, and this means zero extra low frequency distortion, and zero added noise. It also provides a DC blocking form when the music data has a DC offset without the use of a digital DC blocking filter (not easy to do perfectly transparently). It also means that I have eliminated the noise and distortion that the servo adds, because when I remove the DC trim inside the FPGA, I get no change in measurements or sound quality at all.
    I hope this clarifies, Rob
    Chord Electronics Stay updated on Chord Electronics at their sponsor page on Head-Fi.
    https://www.facebook.com/chordelectronics https://twitter.com/chordaudio http://www.chordelectronics.co.uk/
  5. esimms86
    So there's an ADC function sending digital data into the FPGA for eventual conversion to analogue again. Very cool!
  6. mtoc
    Rob, we wanna Dave with WCK ports (both in and out), that will be more suitable with some sources.
  7. bigfatpaulie
    'We' wanna DAVE with WCK ports?  
    I don't want that at all.
    Heck, I don't even know what a "WCK" port is.
  8. Sunya
    Word Clock output
    But since Dave jitter immunity is so good, there's no need for a Word Clock output to synchronize the source to Dave's clock. Plus, there aren't many digital sources with Word Clock inputs.
  9. Rob Watts
    I have produced DAC's and CD players in the past with WC inputs and outputs, and the system worked very well. Then the DAC 64 had a local effectively zero jitter clock and a large buffer, so you could eliminate source jitter.
    But the buffer was useless with video, so in 2000 I started on the path to eliminate the need of the buffer using a digital phase lock loop (DPLL). After about 6 or 7 years I perfected it, and now you can barely hear the difference between DPLL and using a local clock and buffer. Indeed, with what I know about source galvanic isolation, I think it would be a big backward step to use word clocks.
    And of course Dave using USB has full galvanic isolation and has the benefit of clocks being sourced from the DAC too and it works with video!
    Chord Electronics Stay updated on Chord Electronics at their sponsor page on Head-Fi.
    https://www.facebook.com/chordelectronics https://twitter.com/chordaudio http://www.chordelectronics.co.uk/
    onsionsi likes this.
  10. Articnoise

    Yes the more I learn about DAVE the more unorthodox and technical special it seems.


    Roy informed earlier that Devialet use an ADC on its analog input. I thought it was a strange thing to do, but maybe they also are using it with DC coupling caps and to filter out distortion. Probably not the only reason I liked the Devialet better with the SOTA turntable than digital, but it sound really well this way nevertheless. 

  11. ecwl
    I wonder if Articnoise and esimms86 are misunderstanding what Rob Watts is saying about ADC. There's no analog inputs for ADC into DAVE. I think Rob Watts is exclusively addressing the issue of how to handle DC offset and his approach is to avoid using DC blocking capacitors. As a result, he has to have a different way of measuring the DC offset and addressing it. And his approach is to use an ADC to measure the DC offset and then digitally correcting for the DC offset in the FPGA before sending the final digital signal to the pulse array DAC.
  12. esimms86

    I totally get the fact that there are no analog inputs for ADC into DAVE. The fact that there is an ADC function at all is what I find fascinating as further evidence of Rob taking a different approach to solving the digital noise issue with DAVE.
  13. rkt31
    although not owning dave right now but as a future buyer ( whenever i have enough funds ) , i have queries . i use a sub with my kef r300 which are kept close to wall with foam bungs in port. i use rca splitter to feed both sub and power amp from hugo. in case of dave what will be the better option, feeding power amp with rca out and feeding sub with xlr out or vice versa ? my sub is a powered sub . which option will give better transparency ?
  14. Sunya
    The XLRs are 6dB hoter than the RCAs so your sub must have adjustable sensitivity (an input attenuator) to compensate for these differences in output level.
  15. AndrewOld
    Has anyone tried an Intona USB isolator with a DAVE (or TT)? Does it work? Does it offer any benefit? Any thoughts or experiences would be welcome. Seems to me that cleaning up the USB signal can't be a bad idea.
216 217 218 219 220 221 222 223 224 225
227 228 229 230 231 232 233 234 235 236

Share This Page