Passive Volume Control for DACs/Sources

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by Madaboutaudio, Dec 20, 2015.

  1. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    Would you recommend the 10k or 50k version of the khozmo between a Yggdrasil and a tube amp? The 10k has input impedance 10-110k and output impedance 0-100k and the 50k has output impedance 0-500k.
     
  2. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,537
    Trophy Points:
    113
    Location:
    Winnipeg
    Very roughly, you want the pot value somewhere in between the source and amp (log scale), leaning closer towards the amp side if you don't have wiggle room because you don't want to overload the output stage of the dac.

    If your amp has a very low input impedance, then consider a transformer type preamp or something active.
     
  3. OJneg

    OJneg The Most Insufferable

    Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    3,923
    Trophy Points:
    113
    Location:
    Grand Rapids, MI
  4. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    Thank you! This is very useful. Since my tube amp has 100K from input signal to ground, it means that if I have a shunt style attenuator with in-out resistor = 10K and range 0-100K, then the maximum Vout seen at the grid of the first tube I will get = 50/60 Vin = 0.83Vin. This is because I will have two 100K in parallel at the highest attenuator setting.

    With a smaller in-out resistor, Vout : Vin goes up. But then the page says that a larger value minimizes the varying impedance as seen by the signal source. So for the Yggdrasil, what would be a good value for the in-out resistor taking all the above into consideration? 1K? 5K? Or is 10K sufficiently good?

    Thanks!
     
  5. OJneg

    OJneg The Most Insufferable

    Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    3,923
    Trophy Points:
    113
    Location:
    Grand Rapids, MI
    I would save myself a lot of headache and stay away from any questionable shunt attenuators. You can easily get away with a 10k-50k pot. Load seen by the source will be a constant value. Easy peezy
     
  6. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,537
    Trophy Points:
    113
    Location:
    Winnipeg
    Just because I like playing with spreadsheets, I messed around and plotted 3 regular pots 10/50/100k. The deviations from the straight line are due to the parallel load with the amp input impedance (100k) mucking with the voltage division. Overall, not what you would call a big honking deal.

    Granted, this is looking ONLY at impedance values and ignoring whatever witchcraft comes from impedance and capacitance and low pass high pass lube passing cockamamy you want to worry about.
     

    Attached Files:

  7. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    This is awesome, thank you! I followed your lead and created a graph of Delta Vout/Vin for the series pot values 10K/20K/50K/100K assuming the 100K parallel resistor in the amp and 24 linear resistor steps. The results emphasize the differences between the pot values. The ideal is a constant horizontal line. In reality, the pot will be a log pot so the differences will be exaggerated more.

    Given this the 20K pot seems to strike the right balance between high impedance and change close to the ideal. Thoughts?

    @Armaegis Please see if you get the same results with the Deltas to double-check my work.
     

    Attached Files:

  8. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,537
    Trophy Points:
    113
    Location:
    Winnipeg
    Er... did you graph past 100% there? it doesn't make sense for the lines to cross. But yeah, I think you essentially have the same graph that I do. A 10k or 20k pot is just fine.
     
  9. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    This is basically the derivative of your graph so it shouldn't be strange that they cross (I don't go beyond 100%).

    Edit: In your graph, Vout/Vin is the mirror image (using x=y as the reflection line). So then you can see that e.g. 100K goes up slowly and then its rate increases to meet the ideal at 1. The derivative graph shows exactly this behavior. But you're right, it is strange they all cross at one point. The discrete points in my sheet are not exactly the same. I'll look into why there exists an intersection point.
     
    Last edited: May 31, 2016
  10. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,537
    Trophy Points:
    113
    Location:
    Winnipeg
    Eh?? Not sure if we're saying the same thing or not in different words. Such things do not translate well. In any event, just pick a 10 or 20 or be done with it.
     
  11. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    LOL :Violin:
     
  12. uncola

    uncola Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    597
    Trophy Points:
    93
    Location:
    Oahu, Hawaii
    don't forget low capacitance cables between the passive preamp and the amp and as short as possible so there's no high end rolloff
     
  13. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    Thanks! The cables I'm intending the use are 1m long. 1m from DAC to passive, and 1m from passive to amp. They have the following characteristics:
    Z: 55Ω, capacitance: 98pF, R Signal: 0.065Ω, R Ground: 0.008Ω
    What do you think?

    I've decided to order the passive pre from Khozmo with their 25K series pot (they don't have 20K for the series attenuator version). So it'll be this:
    http://www.audiophonics.fr/en/compo...ereo-48-positions-shunt-cms-50k-1-p-8686.html
    inside this: http://www.khozmo.com/stereo_selector.html
     
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,941
    Trophy Points:
    113
    Location:
    Padre Island CC TX

    LOL, the Melos SHA-A Gold head/pre amp had a photoresistor! Anyone who even did the smallest mods on the SHA-1 got rid of it. I'm not sure the reasons, but it may been reliability or lack thereof.

    If you want colorless, I would avoid autoformers and active stages if possible. Listen to @OJneg. I've been through three different pairs of MC transformers and all of them sound colored. The best ones are the least colored, but still colored, in a "good" way. While autoformers (no gain) will probably have less of an effect on the sound than MC transformers, my hunch is that they will not be totally transparent. (There's a Moth autoformer preamp lying around that I should check out sometime.) Same thing with active stages - they will color the sound.

    If you want to the least amount of coloration, go with a quality stepper with custom cables made as short as possible. This is assuming the drive ability of your source is sufficient, otherwise go with an active preamp. We have a Khozmo switcher in the lab and it's really nice. We would probably never offer it because it does need some maintenance in terms of adjusting and tightening up the nut in the back. The electronic resistor ladder switch things that I think @OJneg has is another approach. Bottom line is that if you want transparent to the source, nothing is going to beat two resistors. Go with a 50jk or 100k pot. This is consumer stuff and 50k to 100k will provide you with the best future compatibility.

    That being said, you may want to consider the active or transformer options which have been suggested - they do bring other things to the table.
     
    Last edited: May 31, 2016
  15. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    For kicks, I repurposed an enclosure for something I will never use, put in a new Alps pot I had ordered a while back + some wire, and built a passive.
     

    Attached Files:

  16. OJneg

    OJneg The Most Insufferable

    Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    3,923
    Trophy Points:
    113
    Location:
    Grand Rapids, MI
    oh no! not an Alps RK27! The audiophiles will eat you alive!

    :eek::D;)
     
  17. uncola

    uncola Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    597
    Trophy Points:
    93
    Location:
    Oahu, Hawaii
    omg so common like a fleet street tart charging tuppence a go.
     
  18. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,306
    Trophy Points:
    113
  19. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    At least the signal wires aren't covered in teflon tubing so they won't rattle :D
     
  20. batriq

    batriq Probably has made you smarter

    Pyrate
    Joined:
    Oct 31, 2015
    Likes Received:
    1,041
    Trophy Points:
    93
    OMG, my non-audiophile approved (or audiophile non-approved) passive is f'ing great. More details, more resolution, clearer, more details. Did I mention more details?
     

Share This Page