Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 7 of 7

Thread: DC Coupled CS42448 osh park board

  1. #1
    Junior Member
    Join Date
    Dec 2018
    Posts
    1

    DC Coupled CS42448 osh park board

    Hiya

    I am looking at options for a diy solution for bridging VCV rack to eurorack and the myriad of new affordable analog synths that are cropping up.

    Paul's CS42448 board keeps coming back as one of the coolest and most convenient places to start.

    Has anybody managed to safely DC Couple the ins and outs yet? For synth developers a eurorack specific version of this board will be amazing.

    My electronics skills are still noob level (I'm have way more experience with software development) although I am getting the hang of SMD based eurorack kits and have been studying the O_C module as well as some of the mutable instruments stuff for clues as to how I could do it myself.

    Any pointers would be appreciated.

    PS: Thanks for all the hard work, I am looking forward to getting eyeball deep into teensy development!

  2. #2
    Junior Member
    Join Date
    Jan 2019
    Posts
    7
    Have you had any luck? I'm interested in the same thing.

  3. #3
    Senior Member PaulStoffregen's Avatar
    Join Date
    Nov 2012
    Posts
    20,068
    The CS42448 datasheet shows 2 recommended filters on page 57. Since the goal of this board was just to test the CS42448 chip and get the software side working, I went with the simpler passive filter (figure 31).

    Click image for larger version. 

Name:	filters.png 
Views:	20 
Size:	35.7 KB 
ID:	15788
    (click for full size)

    The more complex active filter (figure 30) looks like it's a proper difference amplifier, so there's a good chance its opamp output will be very close to zero volts when the signal is 0. Maybe it would be close enough if you just replace that 22 uF capacitor with a wire?

  4. #4
    Junior Member
    Join Date
    Jan 2019
    Posts
    7
    Thank you Paul!
    To make sure I understood it right, I have a few questions.
    Adding this active filter after the output of the CS42448 board's DAC, as well as after any DAC like the Audio Shield's one, should provide a DC coupled output?
    If so, how could I use it after the output of the Audio Shield's SGTL5000 since it doesn't have differential audio outputs?
    Which frequencies does this circuit filter?
    Replacing the last 22F capacitor at the far right of the circuit should center the signal at 0V even though there's no bias input?
    By any chance, would removing the 2.2F capacitor at the end of the Audio Shield's line outputs be enough to stabilize the output of DC signals?

  5. #5
    Senior Member PaulStoffregen's Avatar
    Join Date
    Nov 2012
    Posts
    20,068
    Quote Originally Posted by Edouard View Post
    Adding this active filter after the output of the CS42448 board's DAC, as well as after any DAC like the Audio Shield's one, should provide a DC coupled output?
    This circuit is specific to CS42448.


    Replacing the last 22F capacitor at the far right of the circuit should center the signal at 0V even though there's no bias input?
    Removing the capacitor doesn't cause the signal to be any particular way.

    The idea would be to build and test this circuit. Odds are good that output will have 0 volts DC, or very close to zero. Or perhaps some small offset which could be corrected with additional circuity. But this is far from certain. Actual testing is needed.


    By any chance, would removing the 2.2F capacitor at the end of the Audio Shield's line outputs be enough to stabilize the output of DC signals
    This answer is certain: definitely not! The single-ended STGL5000 signals have ~1.5V DC bias. You can easily check this with a DC voltmeter. Just touch the STGL5000-side of the capacitor with your multimeter lead while it's running.

  6. #6
    Senior Member
    Join Date
    Feb 2013
    Posts
    563
    Quote Originally Posted by pixael View Post
    Hiya

    I am looking at options for a diy solution for bridging VCV rack to eurorack and the myriad of new affordable analog synths that are cropping up.

    Paul's CS42448 board keeps coming back as one of the coolest and most convenient places to start.

    Has anybody managed to safely DC Couple the ins and outs yet? For synth developers a eurorack specific version of this board will be amazing.

    My electronics skills are still noob level (I'm have way more experience with software development) although I am getting the hang of SMD based eurorack kits and have been studying the O_C module as well as some of the mutable instruments stuff for clues as to how I could do it myself.
    ... the conditioning circuits i've used for O_C won't be entirely suitable for this kind of thing; they're geared towards precision CV / "quantizing", but you've probably figured that out by now.

    fwiw (i'm not an EE, so take this with a grain of salt), i've made another little thingie (https://github.com/mxmxmx/usb_audio), which is a DC-coupled usb audio interface (using CS4272). to minimize the parts-count, at the outputs (differential to single-ended) i've used (per output signal): 1x INA134 in series with 1/2 of a dual op amp (non-inverting; for scaling up to eurorack levels); at the inputs, another 1/2 dual amp for buffering/attenuation, then 1x THS4521 (single-ended to differential). that worked out fairly ok.

    on a different note: re "bridging to VCV" -- is there even something along the lines of a 6 in / 8 out usb audio object? I'd imagine that's the trickier part to get working ... (?)

  7. #7
    Junior Member
    Join Date
    Jan 2019
    Posts
    7
    Quote Originally Posted by PaulStoffregen View Post
    This answer is certain: definitely not! The single-ended STGL5000 signals have ~1.5V DC bias. You can easily check this with a DC voltmeter. Just touch the STGL5000-side of the capacitor with your multimeter lead while it's running.
    Thank you so much for your explanations Paul! I have forgotten almost everything about electronics since my engineering school days, but I'm learning fast since it's only been a few days and I have a working breadboard (that's far from ready to be produced and used as a Eurorack Module). I suggested removing the capacitor as I first understood that it was here to remove any offset and filter out any DC voltage, but yesterday I read about op amps—had totally forgotten them— and now I understand I will need one with negative feedback after each output to have control on the offset as well as amplify the signal to Eurorack level — I'm thinking of targeting 0V to 8V so that the gate output can be used as an enveloppe. My project has a gate output and an audio output, so I wouldn't need the precision CV out that O&C has and I think I can make it happen with the Audio Shield and its SGTL5000. Time will tell

    I've been inspired to read about op-amps after trying like OP to decipher the much famous and useful Ornament & Crime's circuit diagram, so it's really funny and mysterious that mxmxmx replies right after you. I'm honored to be helped at the beginning of my quest by you two DIY legends. I've read Tom Whitwell say in an article that he could not have done the Turing Machine without the help of many people, which is a bit scary as it shows how hard such a project can be. I've had a hard time finding any documentation to learn about building a Eurorack module, and the closest thing I've found is to read Make Analog Synthesizers by Ray Wilson, do the tutorials for the Teensy and Audio Shield and write my own sketch on the EuroShield— easy for me as once upon a time I was a developer,—adapt it for a breadboard with just the Teensy and the Audio Shield, and finally try to build a breadboard version of the O&C—the closest and best known Teensy-based Eurorack module—to understand how it works and how I can adapt some of its techniques to finalize my breadboard version.

    Maybe having the greatest minds of the Eurorack and DIY community unite to produce a consensus of the best practices in electronics would help to give a good start to newcomers, to avoid some mistakes or incompatibilities in some modules and to encourage the development of even better ways to solve some problems rather than copying circuits coming from PDFs of books written in the 70s. I may be very wrong in all this as I'm a strict newcomer, and maybe trial and error is expected in the electronic world, it's just very surprising when you have a computer science background where everything is binary, true or false, never in between. =)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •