Schiit Yggdrasil V2 upgrade Technical Measurements

Discussion in 'Source Measurements' started by atomicbob, Feb 20, 2018.

  1. atomicbob

    atomicbob dScope Yoda

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    6,126
    Dislikes Received:
    2
    Trophy Points:
    113
    Location:
    On planet
    The V1 and V2 measurements may be compared directly.

    Some explanations.

    The following is called "boiler plate". It precedes the measurement suite and is typical for many measurements. I've become more detailed in my boiler plate as a result of so many questions.

    1. PrismSound dScope III, picoscope 5243B
    2. DAC balanced (Bal) output XLR and unbalanced (SE) output RCA
    3. 100 Kohm load used for measurements
    4. dScope analyzer sample rate 48 KHz unless otherwise noted
    5. DAC 44.1 KHz sample rate, 24 bit depth unless otherwise noted
    6. USB cables – Schiit Pyst and Audioquest Forest
    7. SPDIF input - Tecnec 75 ohm SMPTE 259M/292M Digital Broadcast Cable
    8. Unbalance RCA cable - Worlds Best Cable Gotham GAC-2 or DH Labs Silver Sonic Air Matrix
    9. Balanced XLR cable - Canare L-4E6S starquad with Neutrik XLR connectors
    10. Vaunix Lab Brick USB hub
    11. Shielded 14AWG and 16AWG power cables

    The picoscope 2205 and 5243B may be found on their website. The 5243B is a 16 bit high resolution oscilloscope whereas the 2205 is a more typical 8 bit. Most oscilloscopes are 8 bit with some hitting 10 or 12. I've upped my game for waveform resolution but it has zero effect on the square wave analysis provided for V1 or V2.

    The list of USB cables is there to describe what I use when needed. Both the picoscope 5243B and PrismSound dScope are USB devices. If there are any DAC USB measurements they will also use a cable from the boiler plate description. I have yet to perform USB DAC measurements on the Yggdrasil V2. It is a matter of time. 16 dedicated hours for updates, measurements, data vetting, organizing and posting was all I had available before my vacation begins. Those measurements will occur at a later date.

    Power cords are shielded. Typical EMI screen room technique. Minimizes possibility of measurement contamination from nearby SMPS wall warts.

    Most of the cables listed are those I keep within arm's reach for whatever measurement I am performing. Not all are used on a given DAC during any specific measurement suite.

    Your second question I will refer to the very first statements in my Technical Measurements suite:

    If you are unfamiliar with audio measurements please use a search engine with the query:
    "audio measurements" or "audio measurement handbook"
    Look for publications by Richard C. Cabot and also by Bob Metzler, both from Audio Precision. There are other useful publications as well. These will provide basic knowledge.
    Interpretation of the following measurements is beyond the scope of this post.

    I don't wish to get into data interpretation, at least not until after I am retired and have more time to address such questions.

    Here is a reference you may wish to read:
    http://www.nanophon.com/audio/diagnose.pdf
     
    Last edited: Feb 20, 2018
    Merc, 9suns, FutureFrenzy and 3 others like this.
  2. atomicbob

    atomicbob dScope Yoda

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    6,126
    Dislikes Received:
    2
    Trophy Points:
    113
    Location:
    On planet
    It's on the todo list for later, after vacation. That is if you mean physical thermal warmup. I won't be tracking jitter during temperature measurements. While interesting, that is very complicated and takes the equipment out of service for other measurement duty for possibly up to a week.
     
    Dino likes this.
  3. Cspirou

    Cspirou They call me Sparky

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,941
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Irvine,CA
    Yeah, best to save it until electric measurements are completed.

    Is testing with a 600 ohm load in the measurement queue as well? Its one of the features Schiit mentions for the upgrade and was a shortcoming in the Stereophile review.
     
    A1Gear likes this.
  4. bobsherman

    bobsherman Acquaintance

    Joined:
    May 14, 2016
    Likes Received:
    40
    Dislikes Received:
    0
    Trophy Points:
    18
    Same DAC different person doing the testing. One (Bob) is an expert and one (the amir) is an, _ _ _hole! fill in he blank!
     
    Elnrik likes this.
  5. bobsherman

    bobsherman Acquaintance

    Joined:
    May 14, 2016
    Likes Received:
    40
    Dislikes Received:
    0
    Trophy Points:
    18
    The amir says his is not calibrated, as it would cost too much money to have it done. He figures any deviations would be consistent. He is a real scientist. Not!

    He is just a hack with an expensive toy.
     
    Last edited: Feb 20, 2018
    earnmyturns likes this.
  6. Elnrik

    Elnrik Super Friendly

    Friend
    Joined:
    Jan 10, 2017
    Likes Received:
    3,658
    Dislikes Received:
    17
    Trophy Points:
    113
    Location:
    Denver CO
    Enjoy your vacation @atomicbob !

    I know you deserve it!
     
    Last edited: Feb 20, 2018
  7. johnjen

    johnjen Friend

    Friend
    Joined:
    Sep 29, 2015
    Likes Received:
    522
    Dislikes Received:
    2
    Trophy Points:
    93
    Location:
    Well Grounded
    I am enthralled with the differences that the v.2 boards make, especially in the square wave response at 20Hz. It now much more closely matches the response that the PWD mkII (my secondary dac) delivers and I'm hoping it will bridge the gap in the audible performance between them as well.

    And the reported improvements in the entire mid range on up, as exemplified by the 'limbo' measurements (1KHz @ -145dB, wholey schlamolie…!) will be very welcomed as well.

    Now if only Schiit would speed up their updating capacity. It looks like they have one guy doing the upgrades which means they average about 1 upgrade per day and with me being in the cue at ≈120 that's 4 more months until my number rolls around.

    I'm REALLY glad I don't suffer from the persistent and seemingly ubiquitous Audio Nervousa syndrome, but then just having a Jggy in the first place goes a long way it settling those Nervousa twitches and whole body spasmodic contractions.
    hahahahahahahahahahaha

    JJ
     
  8. Darren G

    Darren G Friend

    Friend
    Joined:
    Apr 4, 2017
    Likes Received:
    490
    Dislikes Received:
    8
    Trophy Points:
    63
    Location:
    Pennsylvania, PA
    If true, it will be 5 months before my Yggy is upgraded, but I think we are better off with Schiit doing sane business (too many small/medium businesses drive themselves into the ground by being over reactive to the 1-3% of their core business) .

    That's also 4-5 months we can enjoy the Yggy's we have in the mean time ;)
     
  9. Ash1412

    Ash1412 Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    264
    Dislikes Received:
    2
    Trophy Points:
    63
    This huge upgrade in linearity and the limbo test leads me to believe they changed the analog board topology to something closer to the soekris dacs. Isn't 144 db equivalent to 24 bit performance? How the heck did they do this with only "obsolete" 20 bit chips? Amazing work by schiit! Wonder what stereophile would say about yggy v2
     
    Alcophone likes this.
  10. Cspirou

    Cspirou They call me Sparky

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    2,941
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Irvine,CA
    I wouldn't say it's actually resolving down to 144dB. The 120Hz linear switching noise starts to takeover at that point. But there is a clear signal for sure and definitely resolving to 21 bits as Schiit has claimed. Very impressive
     
    Darren G likes this.
  11. Ash1412

    Ash1412 Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    264
    Dislikes Received:
    2
    Trophy Points:
    63
    What i meant is that the DAC section is capable of outputting a 1k sine 144db down/24 bits, regardless of whether that is masked by the ps noise, masked by thermal noise or is even audible. How Mike did it using 20 bit chips is pure magic. On the subject of ps noise, since the Yggy already has a pretty huge shunt regulated power supply, this might be the lowest we can practically ever get.

     
    Last edited: Feb 21, 2018
  12. Clemmaster

    Clemmaster Friend

    Friend
    Joined:
    Sep 28, 2015
    Likes Received:
    660
    Dislikes Received:
    0
    Trophy Points:
    93
    It's not magic. It's called dithering and is the basis of noise shaping as used by Sigma-Delta cheaps.

    It's also used in most consumer LCD monitors: while they claim to output 8-bit colors, their panel is often 6 bit and they make up for the difference using FRC, which is a form of dithering.
     
    Telstar and Darren G like this.
  13. Andre Y

    Andre Y Friend

    Friend
    Joined:
    Oct 4, 2015
    Likes Received:
    157
    Dislikes Received:
    0
    Trophy Points:
    43
    Location:
    Southern California
    You have to be careful when you look at the noise floor to also consider the size of the FFT bins used to analyze that signal. There is a certain amount of noise in the measurement bandwidth (say 22.05 kHz for 44.1 kHz sampling). When you add more bins for the same bandwidth, the noise floor appears to go down because you're dividing the same amount of noise over a larger number of bins. This isn't fake (the noise floor is actually decreasing per bin), but it may not be relevant either depending on a number of factors. To calculate SNR, you need to add up all of the noise in the measurement and use that instead of looking at the noise floor at any one point.
     
    Priidik, murray and Cspirou like this.
  14. Ash1412

    Ash1412 Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    264
    Dislikes Received:
    2
    Trophy Points:
    63
    Wouldn't they have to implement that on the DSP chip though? I'm inclined to believe they didn't alter anything besides the analog board. Mike also doesn't seem to support dithering from some of his earlier hf posts
     
  15. earnmyturns

    earnmyturns Friend

    Friend
    Joined:
    Sep 25, 2016
    Likes Received:
    862
    Dislikes Received:
    0
    Trophy Points:
    93
    Location:
    Palo Alto
    Home Page:
    Schiit has said that the Analog 2 upgrade requires firmware changes. That firmware is where the mega-combo-burrito filter/upsampler lives. What new magic @baldr put there, only he and his small DSP posse know.
     
    Torq likes this.
  16. Torq

    Torq Money can't buy class

    Friend
    Joined:
    Jan 15, 2016
    Likes Received:
    7,887
    Dislikes Received:
    11
    Trophy Points:
    113
    Location:
    Seattle, WA
    The DSP chip is a programmable part. It runs a program that's loaded from a standard ROM chip (the little socketed 8-pin DIP chip you'll see on the DSP board ... clearly visible in their product shots ... behind the USB input board). Depending on when your Yggdrasil was last at Schiit it may have the original program on that chip or it may have had it's ROM replaced.

    If you send in a unit that has the original DSP code on it, they'll just replace that chip with one that has the updated code on it. 'Tis why it is in a socket.
     
    Darren G and A1Gear like this.
  17. monacelli

    monacelli Friend

    Friend
    Joined:
    Jul 23, 2017
    Likes Received:
    497
    Dislikes Received:
    0
    Trophy Points:
    63
    Location:
    PNW
    Mike answered a related question on Head-Fi about getting greater than 20 bits of resolution out of 20-bit chips. Here's a link: HF Yggdrasil Impressions.

    And the relevant Q&A:

    "Q: Thanks! And how does it end being a 21 bit DAC with one 20 bit chip per phase per channel? Not important, just curious.

    A: Two twenty bit DACs (one per phase) double the resolution for a balanced signal. Double the resolution only adds 6db (one more bit) for a total of 21 bits. In this case, 20+20 equal 21."

    So I think the answer is a combination of the above and the bin size considerations mentioned by @Andre Y.
     
    Last edited: Mar 9, 2018
    slankoe likes this.
  18. Telstar

    Telstar Bottom 1% of posters on SBAF

    Joined:
    Mar 28, 2018
    Likes Received:
    6
    Dislikes Received:
    12
    Trophy Points:
    8
    Location:
    Italy
    These measures look really good.
    Hope to see a Metrum with dac2 modules measured here soon.
     
    Don Quichotte likes this.

Share This Page