Modi Multibit i2s input mod

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by Scott Kramer, Nov 8, 2016.

  1. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Don't hold your breath... lot's a fun stuff in the mean time to work on / play with ;)

    ^ *******

    NEW Thread! Cool... this was a bit of a teaser I put in the PI threads, when @Michael Kelly was mia!
     
    Last edited: Dec 14, 2016
    meloman likes this.
  2. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Usually when it's quiet's when the cool stuff's going on :D

    ...But I'm in the dark re: @Michael Kelly

    [​IMG]
     
    Last edited: Nov 10, 2016
    meloman, Wfojas and pandather like this.
  3. pandather

    pandather Acquaintance

    Joined:
    Feb 8, 2016
    Likes Received:
    65
    Dislikes Received:
    0
    Trophy Points:
    18
    Is it just me @Scott Kramer or do I spy i2s directly into a Modi Multibit?!?
     
    Scott Kramer likes this.
  4. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Yes...

    I need to start a separate thread, have real music playing just tonight, not white noise, so kinda excited about it! -->


    This is my Schiit i2s "possible or not" study... to see if there's mysterious "show-stopping" logic between the USB chip / (COAX-OPT) chip -- DSP chip -- DAC chip. This was my research, basically tracing the board, looking up chip pinouts, measuring voltages, then doing the minimum, least traumatic, physical hookup & see if it's a *go* or *no go*. Plus it's fun to see how they pulled off this little bad boy.

    (I chose to cut off USB by the balls)...


    This is just the initial test. i2s is meant for short board traces so It will need precise U.FL connections, a correct resistor array (measured with a good oscilloscope) etc. Ha... was kinda hoping for a solid no-go :D

    On a bit of a quest after hearing what that Kali FPGA FIFO i2s clocker can do... it's trapped behind standard raspberry pi Dac's--!!! (And would like to implement direct, unencumbered i2s in general)

    PS. being fairly carefully doing this... so I can go back to stock.

    [​IMG]
    [​IMG]
    [​IMG]
     
    Last edited: Aug 22, 2018
    ultrabike, bazelio, meloman and 10 others like this.
  5. Michael Kelly

    Michael Kelly MOT: Pi 2 Design

    Friend
    Joined:
    Apr 22, 2016
    Likes Received:
    1,883
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    Rhode Island, USA
    Home Page:
    Wow - very crazy and very awesome! I have the CM6331A datasheet, but I can't copy it (under NDA). I could however share the pins where the I2S is located if you'd like!

    BTW, the 502DAC has a header for the I2S. Because we use the PCM5122A in master mode, it generates the MCLK from the selected low noise NDK oscillator. You could easily hook another downstream DAC to this header.

    Cheers,
    Michael

    P.S. 6 502DAC's will ship out tomorrow. I'm still having trouble choosing who gets them since I had hoped to have 10, not just 6. I plan to build another 6 next week though, so hopefully whoever is left out this pass won't have to wait too long.
     
  6. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Hi @Michael Kelly! Yep just a bit crazy, fun to see how they built it :D. Pretty straight forward once you trace through it all. They flip sides a lot thorough tiny vias, it's a maze!

    I did find the CM6631A's i2s out, removed a quad resistor array (used for each of the i2s lines) to take it out of circuit... Thanks!

    Next task is to tell its DSP to keep accepting the i2s signal when the USB connector is out... already found some single pins changing state wether plugged in or out, so probably doable.

    Good news on the 502DAC's.
     
    Last edited: Dec 19, 2016
  7. IceUul

    IceUul Friend

    Friend
    Joined:
    Oct 20, 2015
    Likes Received:
    100
    Dislikes Received:
    3
    Trophy Points:
    43
    Location:
    Estonia
    @Scott Kramer

    Did direct i2s made any difference? I mean does it sound better now?
     
  8. Michael Kelly

    Michael Kelly MOT: Pi 2 Design

    Friend
    Joined:
    Apr 22, 2016
    Likes Received:
    1,883
    Dislikes Received:
    1
    Trophy Points:
    93
    Location:
    Rhode Island, USA
    Home Page:
    They are likely simply detecting the presence of the 5V on the USB power pin. If you can tie 5V to that pin it should be happy.

    Cheers,
    Michael
     
  9. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    It's interesting, but not really listening/comparing yet. Cleaning up and simplifying the connection now, I'll see where it stands a bit after that.

    Cool, I'll look for that--
     
  10. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    @Michael Kelly Thanks for the awesome hint... I noticed the USB is powered even when the DAC is unplugged from the wall, you can tell because it still shows as an audio device in macOS. Made it even easier to trace. First the USB 5V goes to a 3.3V regulator... then that voltage goes to the DSP pin as 3.3V when USB plugged in... simply jumped that pin to the moby's always on 3.3V. 10min job!

    how 'bout a vid instead of a wall -o- text :D -->



    FYI @Michael Kelly did it that way 'cause didn't want to put the 5V directly on the USB & possibly interfere when I plug in USB, plus the 5V was far away & separate from what I could tell--
     
    Last edited: Apr 2, 2017
  11. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    I just had to add features :D and getting ready to button it up-- need advice on the case.

    Using a 2x5 pin header to give access to *both* usb i2s out, and i2s in--> that way you can use jumpers to tie the USB back in (basically returning it to stock operation).

    pics:

    i2s mode:

    [​IMG]

    usb mode (reenabled by adding standard jumpers): btw, this adds an independent usb i2s out, that can run alone or when using coax/optical -- may be hit or miss on the usb-->i2s out, they used a large value 1.2Kohm resistor array, this is very large, probably needed for the dsp, I reused this resistor array.

    [​IMG]

    analyzing/debugging i2s:

    [​IMG]

    here's my dilemma, how to mod the case... like I marked here, or do you guys have a better idea: (also any steel cutting tips, good dremel tip to use, etc)

    [​IMG]
     
    Last edited: Dec 5, 2016
  12. pandather

    pandather Acquaintance

    Joined:
    Feb 8, 2016
    Likes Received:
    65
    Dislikes Received:
    0
    Trophy Points:
    18
    Scott, if you made a guide to mod it how you did, I would definitely be interested! That's some awesome work by the way. I wonder if a small DIP switch could be used to toggle i2s or USB in, and whether or not to allow USB to i2s otherwise. :)
     
  13. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Listened to the i2s moby on the big system last night, well... didn't 'f it up at least. It was awesome, meat on the bones as usual yet so freaking fast, 100's of shades of bass, the phasy things come forward-- and big soundstage at low volume... pushed 3am. Need to go back to spdif and see if it changes, expectation bias in play of course ( I'll need a month :D ) ---- sorry, really have a hard time talking about sound quality and differences.

    ...and was just the touch, not even the pi/kali yet.

    [​IMG]
     
    Last edited: Dec 8, 2016
  14. Merrick

    Merrick A lidless ear

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    5,950
    Dislikes Received:
    13
    Trophy Points:
    113
    Location:
    Portland, OR
  15. zerodeefex

    zerodeefex Grumpiest admin

    Staff Member Friend BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    12,248
    Dislikes Received:
    57
    Trophy Points:
    113
    Zomg you beautiful genius. This is ridiculous and kind of amazing.
     
  16. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Friend MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    11,354
    Dislikes Received:
    40
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    @Scott Kramer My transport does I2S out via ethernet, and I would absolutely love to hear the Schiit MB DACs hooked up like that, even if just for shits and giggles or insanity.
     
    Scott Kramer and Dino like this.
  17. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    Was thinking Yggdrasil, much more space to work in anyway ;) kidding (a little), need to see where they're going with that "digital" box hinted at and next Yggdrasil upgrades.

    Haha :D genius is quite a stretch, really like ridiculous & beautiful tho, lol! Hope you read thru my reasons in the pi threads. Thanks!

    Cool! Would probably be easy to wire that in.


    Guys, I'll have more detail in the pi threads-- added a few more features.

    At this point it's solid and looks good, seems to sound incredible, and at a great stopping point.


    ********** Warning......

    *** More importantly, there's reasons *** not *** to do this (well, I'd still do it ;)), specifically they may be communicating sample rate to the DSP outside just the i2s. Right now I have to upsample/downsample/keep native --> everything to 24/96. This is not clear yet, it may be an i2s problem from squeezelite, or my compile of linux (bleeding edge 384K kernel). To me, no big deal, LMS server is awesome with it's rules and resampling on the fly with sox. But no doubt a show stopper for some.

    ...ever hear the moby "wig" out? in normal use it happened to me once or twice a month, where there's high frequency "crickets" glommed onto the signal?
    ...that's what happens feeding it something other than 24/96 thru i2s right now, with my sources.

    **********

    So at this point, if I discover whats going on re: i2s, and is coming from *only* the i2s, may take it to the conclusion, wire it with U'FL (mini-coax) and measure the perfect resistor array when using the kali FPGA clocker (this tunes the square waveforms, you get overshoot/undershoot if not right (impedance mismatch/ringing), easy to see on a scope)
    -- anyway from what I'm learning, I'm no expert.

    Thanks!

    PS. Some interesting SPDIF in the other pi thread, just post today.
     
    Last edited: Dec 16, 2016
    Kattefjaes, HitmanFluffy and Dino like this.
  18. Scott Kramer

    Scott Kramer Friend

    Friend
    Joined:
    May 3, 2016
    Likes Received:
    1,272
    Dislikes Received:
    3
    Trophy Points:
    93
    oops, wrong button
     
  19. Cspirou

    Cspirou They call me Sparky

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    5,716
    Dislikes Received:
    5
    Trophy Points:
    113
    Location:
    Northwest France
  20. landroni

    landroni Friend

    Friend
    Joined:
    Aug 12, 2016
    Likes Received:
    1,295
    Dislikes Received:
    8
    Trophy Points:
    93
    Yeah, some of us seem to be marked these days... :) When's the next sacrifices to the SBAF audio gods happening?
     

Share This Page