atomicbob's SMPS Noise Nuke for HP amps with external SMPS

Discussion in 'Headphone Amplifier Measurements' started by atomicbob, Sep 26, 2017.

  1. Justin S

    Justin S Friend

    Pyrate Contributor
    Joined:
    Mar 14, 2018
    Likes Received:
    1,506
    Trophy Points:
    93
    Location:
    Toronto
    I am about to build up a couple of these. I found the Hammond 155B chokes at PCX here in Ontario. They have many in stock (I picked up 3 figuring I would make a few of these for a couple PI2AESs and a little SBC I am going to use as my Roon server.

    I have never worked with a choke before. Is there a correct flow from feed to drain? The two wires are not labeled and the spec sheets do not indicate which wire is which. I would be grateful for any help with this.
     
  2. RestoredSparda

    RestoredSparda Friend

    Pyrate
    Joined:
    Jan 7, 2021
    Likes Received:
    3,634
    Trophy Points:
    93
    Location:
    WI
    Nope. Orientation does not matter.
     
    • Agreed, ditto, +1 Agreed, ditto, +1 x 2
    • Like Like x 1
    • List
    Last edited: Sep 28, 2022
  3. Justin S

    Justin S Friend

    Pyrate Contributor
    Joined:
    Mar 14, 2018
    Likes Received:
    1,506
    Trophy Points:
    93
    Location:
    Toronto
    On more quick question. My barrel connectors are metal and will connect the input and output jack (-) sleeves to the case. I know that Atomic Bob's circuit grounds to the case at a single point off of the (-) side of the caps. Will having both input and output jack (-) sleeves connected to the case cause issues with the circuit or is this the same as wiring the (-) side to the case?

    EDIT: Looking at the build by @Azimuth , it looks like having jacks that ground to the case is ok.
     
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
    Last edited: Oct 3, 2022
  4. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,975
    Trophy Points:
    113
    Location:
    Vancouver BC
    Yep, the input and output -ves are connected to each other and grounded to the case: see @atomicbob 's circuit diagram in this post, and the pics of others' builds.

    Just make sure to label the jack you connect the caps to as output so you don't do a @RestoredSparda ;)
     
    • Like Like x 2
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
  5. Justin S

    Justin S Friend

    Pyrate Contributor
    Joined:
    Mar 14, 2018
    Likes Received:
    1,506
    Trophy Points:
    93
    Location:
    Toronto
    Thanks @Biodegraded . I am not great at reading schematics and tend to look at them rather literally. I appreciate your confirmation here. I did see @Azimuth 's build and it looks very similar to my plan. I am also going to put a toggle switch on mine between the + in and the choke. I'll post photos. I am building 3 of them.
     
    Last edited: Oct 4, 2022
  6. Justin S

    Justin S Friend

    Pyrate Contributor
    Joined:
    Mar 14, 2018
    Likes Received:
    1,506
    Trophy Points:
    93
    Location:
    Toronto
    Did the first part of the build today - drilling, mounting, tinning and connecting the caps together and to the choke. I think there's another afternoon to do on these and i'll be finished. I know, I'm not going to win any awards for my soldering - I end up worrying about mechanical failures so I go a bit heavy.

    I used bolts to mount the choke and Sugaru to mount the caps.

    Looking forward to trying these out when they're done.
     

    Attached Files:

  7. Justin S

    Justin S Friend

    Pyrate Contributor
    Joined:
    Mar 14, 2018
    Likes Received:
    1,506
    Trophy Points:
    93
    Location:
    Toronto
    I got lock-jaw on these and finished them up after dinner this evening. I tested for shorts and continuity, and everything checked out. If I were to do it again, I would pick a slightly larger box - the caps with the zip-tie required me to squish the cases a bit to get them to close.

    I am listening to one of them right now.

    CHAIN: MacMini Roon Server>LAN>Pi2AES Lite (with Allo Nirvana SMPS+ABNN)>Lynx Hilo>BHA-1>HD800S.

    The ABNN should not make a difference in this chain.

    Things sound subtly different with the ABNN.

    With ABNN:

    - the background is blacker
    - everything is more "clear" sounding
    - Trailing reverb, room cues, micro details are more apparent.
    - WTF! Bass seems more full with the ABNN. This could be an effect of the background/clarity improvement.
    - I'm not sure about this one, but I think everything is a bit closer with the ABNN. The staging seems a bit less deep, but there is still great separation. Frankly, with the 800s, I appreciate the vocals being a bit closer.

    Without ABNN:

    - Slightly deeper staging.
    - voices and more recessed and less "wet" sounding
    - the lighter presentation feels "crisp" and very nimble
    - everything is a little more "angular"

    Caveat on my observations - I had to reboot the PI2AES each time I switched to and from the ABNN - the A/Bs were not quick. In addition, my positive observations could be "DIY Pride Bias". I will say I ran through three songs a number of times in both focused "sign posted" mode and "listening while doing other stuff" mode. In each case, I found the ABNN makes my chain more involving to me.

    My initial impressions of the ABNN are quite positive. I am happy that I built a few - a couple for PI2AESs and one for miscellaneous duties. I would like to try this with an analog device like an amp. I think this would be a worthwhile endeavour for anyone who is DIY-inclined. I am not very experienced with electronics, but I found the process pretty easy and a fun way to spend some hobby-time.

    Thanks to @Biodegraded and @RestoredSparda for their input - it was very helpful. Thanks to @Azimuth for his photo documentation which helped me figure out how to translate the schematic into a build. Of course, a big thank-you to @atomicbob for sharing this valuable, practical and well-documented design with the community. ABNN: Nuking not only noise but also nervosa. Bring on the stock PSUs.

    PXL_20221017_010042444.jpg
    PXL_20221017_015104850.jpg PXL_20221017_015054409.jpg PXL_20221017_015136723.jpg
     
    • Like Like x 6
    • Epic Epic x 2
    • List
    Last edited: Oct 16, 2022
  8. penguins

    penguins Friend, formerly known as fp627

    Pyrate
    Joined:
    Sep 26, 2018
    Likes Received:
    3,451
    Trophy Points:
    113
    Location:
    SouthernCA
    Not planning on building one due to a lack of need / equipment running on a SMPS, but just out of curiosity - has anyone built one of these but with more smaller capacitors in parallel? If so, what was the audible / practical difference, if any, vs using 2 bigger caps?
     
  9. Cspirou

    Cspirou They call me Sparky

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    8,200
    Trophy Points:
    113
    Location:
    Northwest France
    You mean the smaller capacitors would have the same total capacitance? Shouldn't really affect much

    If you are looking for a smaller filter, I came across this at diyaudio

    https://diyaudiostore.com/collections/power-supplies-accessories/products/smps-dc-filter-p089zb-kit
     
  10. Aklegal

    Aklegal Friend

    Pyrate Contributor
    Joined:
    Mar 1, 2016
    Likes Received:
    623
    Trophy Points:
    93
    Location:
    Atlanta, GA
    Can we pin this thread to the top somewhere? I just stumbled upon this. I had no idea this existed and I visit this site multiple times a week. I have pretty much scrapped my plans to build an AMB Sigma 11 for my Pi2AES after reading this thread.
     
  11. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    More paralleled capacitors of lower capacitance will be fine as long as total capacitance maintains equivalence and working voltage rating is the same or higher than those specified in the OP.
     
    • Like Like x 2
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
  12. penguins

    penguins Friend, formerly known as fp627

    Pyrate
    Joined:
    Sep 26, 2018
    Likes Received:
    3,451
    Trophy Points:
    113
    Location:
    SouthernCA
    ^ Not to be snarky or weird, apologies in advance if it is, but just to clarify, the question was only asking in terms of sonic differences. I understand the working voltage rating + same equivalent capacitance value means "should work".

    Asking because I was playing around with a DIY amp in the more distant past that needed "noise nuking" and much to my surprise, at one point, I split the L/R channel into 2 of the same circuit and found that for some reason, one channel sounded better and had less noise with 1 big +1 bypass caps while the other channel was better / less noise with 4 smaller caps that added up to the same value, even when I swapped some of the components between the 2 channels (i.e. check components). Messed around with other slightly different configurations of the circuit and another time, it was best with 2 medium caps + 1 bypass. etc.

    At first I thought it was either overlooked human error or maybe had something to do with the RC values picking up or letting through certain / different noise, maybe a phase shift related something in certain frequencies (ex: some songs with inverse phase on Yggdrasil sounds weird but I'm usually not sensitive enough to it to be able to place my finger on it unless someone tells me), or some other circuit theory related stuff. At the same time, while I haven't looked very hard, I haven't found any one individual thing in circuit simulations that always 1:1 directly correlates to a particular aspect of the sound that I hear.

    So, I'm wondering if other listening impressions with different noise nuke configs would correlate to anything I heard with the above. Only problem here is my memory of the exact sonic differences between the different capacitor configs is a bit fuzzy now and I can't find the notes... oops.
     
    Last edited: Jan 29, 2023
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    It may make a sonic difference, maybe. Moar many caps that equal up to same capacitance would result in lower ESR. Another audiophile tweak is higher voltage but same value capacitance. Yet another is paralleling smaller higher quality caps (film).

    Key is different, but not necessarily better. BTW, We can see the moar many caps method taken to the extreme in Denapfrips DACs.
     
  14. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    I remember reading a log time ago... Something something different caps also have different discharge speeds... something something which allows them to potentially more effectively cover a wider band of noise/distortion? Or something along the lines of being more effective at nuking particular frequency spots better than others? Kinda makes sense in my head but I lack the EE verbiage to explain it properly.
     
  15. Pancakes

    Pancakes Friend

    Pyrate Contributor
    Joined:
    Aug 13, 2020
    Likes Received:
    1,417
    Trophy Points:
    93
    Location:
    Atl
    Smaller caps charge/discharge faster than bigger caps. It's purely a size thing. That's why large caps are often bypassed with small caps. As in, you have plenty of energy storage in the big cap but you add a small one to quicken the energy discharge.
     
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
  16. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    @purr1n gave the answer I would if he hadn't beat me to it. Bypass the electrolytic capacitors with some 0.1uF capacitors. For additional nervosa consider the differences between ceramic, mica, mylar, polycarbonate, polystyrene, polyester, metal film etc.
     
  17. lehmanhill

    lehmanhill Almost "Made"

    Contributor
    Joined:
    May 3, 2018
    Likes Received:
    545
    Trophy Points:
    93
    Location:
    Home of Jiffy Mix
    You asked about sonic difference and I can offer on experience and @purr1n's comment on lower ESR for smaller value caps in parallel seems to be the reason. Some years ago, I was working with a DIY power amp that had dual diode bridges and 10,000 uF power supply caps on board. As a comparison, I tried the normal Nichicon caps that had an ESR of about 45 mOhms. Then I tried a pricey Mundorf cap that had an ESR of 6 mOhms. I couldn't tell any difference in noise, but the Mundorf had better bass, tight and clean. Later, I rigged up an offboard variation with smaller normal caps in parallel that was about the same total capacitance (3 x 3300 uF, if I recall). That setup sounded the same as the Mundorf.

    My theory is that the bass notes demanded a high current transient and the CR filter of the normal 10,000 uF formed by the ESR and the capacitance wasn't fast enough. I think that small film bypass caps is a good thing, but the film cap doesn't have the size to supply current to the bass note. To me, paralleling smaller caps + a bypass cap is the best solution. In fact, I'm looking at a CRC power supply on my bench that has 2 10k uF in parallel for the first "C" and 10 x 3300 uF caps in parallel for the second "C".

    All of this to say that the atomicbob noise nuke is a great idea as presented. If you happen to have smaller caps to parallel it certainly won't hurt anything and it could even help a little. I am reminded of the work done on the design of the superregulator in the 90's which showed that noise + low output impedance over the frequency spectrum was correlated to best sounding power supply.

    No longer available, but Jan Didden's SilentSwitcher showed how to achieve low noise and low impedance from a switcher as shown by the curves that I borrowed from the DIYaudio Store information on the SilentSwitcher. silenswitcher B.jpg silentswitcher.jpg
     
  18. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,833
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    I can't find the video, but I remember watching one on a Commodore 64, they tried many types of bypass caps on the power rails to reduce the SMPS noise that would sometimes interrupt the main computer circuit, and the one type that did the best was ceramic.

    I think they even found that the newer types of Aluminum Organic Capacitors did far better than normal Aluminum Electrolytics and with those, a ceramic cap was hardly needed. I have noticed a few of these newer caps on even some Schiit designs in key places. They are not on this cart, as those are severely limited by voltage as you go up in capacitance. However, still a decent chart.

    Capacitor-types.png
     
  19. lehmanhill

    lehmanhill Almost "Made"

    Contributor
    Joined:
    May 3, 2018
    Likes Received:
    545
    Trophy Points:
    93
    Location:
    Home of Jiffy Mix
    I am surprised to see tantalum caps listed. Solid tanatalum caps typically fail in short circuit which is scary enough that I went through my parts drawers and threw out all tanatlum caps. Their other disadvantage for audio is that they have high impedance at high frequency.

    I wonder what they mean by power caps?
     
  20. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Whatever people end up doing with filtering caps, can we at least all agree that Denafrips are nuts?

    [​IMG]
     
    • Like Like x 5
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List

Share This Page