D-Noizator: a magic active noise canceller to retrofit & upgrade any 317-based V.Reg.

Here is my final solution. Comments about the board noted and addressed.

Secondaries can be customised for parallel (A+B, C+D) or series (B+C). I've experimented with custom primary jumpers, but it looks rather messy, so this is 230V for now. Would be responsive if people are really desperate for 110V.

attachment.php


Has been really fun and an enjoyable learning experience. Thanks Nisbeth for letting me loose on it :)

Gerbers attached, UNTESTED, but verified error-free in Eagle.

For some reason, the polygon pads look filled on the copper layers in most free online Gerber viewers but are correctly isolated on PCBway? Not sure why. Going to take this to fabrication in a couple of days.
 

Attachments

  • AC-LM317-V2.0-Board.png
    AC-LM317-V2.0-Board.png
    135 KB · Views: 985
  • AC-LM317-V2.0-Top.png
    AC-LM317-V2.0-Top.png
    242.4 KB · Views: 473
  • AC-LM317-V2.0-Bottom.png
    AC-LM317-V2.0-Bottom.png
    151.2 KB · Views: 512
  • AC-LM317-V2.0-Schematic.png
    AC-LM317-V2.0-Schematic.png
    40.4 KB · Views: 524
  • AC-PSU-V2.0.zip
    132.3 KB · Views: 74
Last edited:
Looks very good.
May I ask why will you order from PCBWay?
I’ve recently switched to JLCPCB. Your design is 40$ at PCBWay and 7$ at JLCPCB. I’ve ordered mine from JLCPCB and quality is excellent. Both ship with DHL Express at same price.

I'll probably use JLC. I bought your board from there, awaiting shipping.

Just wanted to check the Gerbers through different places. Only PCB Way showed the isolated pads.

Otherwise JLC, pcbxprt.com, gerber-viewer.com - they all showed filled pads on the copper layer. Not really sure why.
 
Member
Joined 2019
Paid Member
I did check wih KiCad gerber viewer and it shows as OK. There are some warnings of unexpected characters on job file, but that could be some difference between Eagle and KiCad.
If you don’t use latest Eagle version, it could be good to upgrade (if possible) and check again.
 

Attachments

  • gtl.JPG
    gtl.JPG
    81.8 KB · Views: 136
Yet another PCB layout of the De-Noiser. This one uses a four layer board (!!) and includes a jumper option to disable/enable the De-Noiser. Gerber files freely downloadable.

VRDN: bipolar regulator PCB for line level ckts: ±11V to ±20V @ 1.5A with "De-Noiser"
Thanks for your contribution to the community.

Your version is "maximalized", as noted by other members, but that is globally the trend followed since the thread was started.

Initially, what I had in mind when designing the project was a tiny (probably SMD) 3-wire add-on module that could convert any "normal" 317 regulator into a superreg (kind of), cheaply and easily.

Things turned out differently, and although I am pretty sure that a number of members did actually follow the initial scope without telling anyone, most of the projects posted so far are complete, superreg-like regulators.

I do not have any objection to that, and once a project is posted, it gets a life of its own and its creator doesn't control it any more.

I am certain that the convenience you added by providing a PCB, BOM and other essential aspects will please many members whilst irritating others, but there is room enough for everyone: basically, do as you like and follow your own path.
Since you have started your dedicated thread, the "purists" here won't be inconvenienced by it (but personally, I don't mind).
 
This a builder's place, so it's far from me to criticize how my pals here prefer to build their stuff.

When I commented on how things were diverting from what I thought was the original intention, improving an existing regulator, it was just to point it out.

I keep working in that direction. In fact I will be soon uploading another "prototype board", this time from Diego's Dienoiser.

I will also upload another version of the Denoisator, which had some mistakes in it.
 
This a builder's place, so it's far from me to criticize how my pals here prefer to build their stuff.

When I commented on how things were diverting from what I thought was the original intention, improving an existing regulator, it was just to point it out.

I keep working in that direction. In fact I will be soon uploading another "prototype board", this time from Diego's Dienoiser.

I will also upload another version of the Denoisator, which had some mistakes in it.

Thx for your effort and yes I think it is quite fun and learning from this as my great interest in electronics and hifi😊
 
Initially, what I had in mind when designing the project was a tiny (probably SMD) 3-wire add-on module that could convert any "normal" 317 regulator into a superreg (kind of), cheaply and easily.

I guess this layout is more according to Elvee's original vision :D
Electrolytics are Nichicon low-profile (100u/35V and 22u/35V).
It is a dienoiser with denoiser stuff in the bottom and dienoiser add-ons on top. However dienoiser seems to be very prone to oscillations and I could not measure any real improvements agains denoiser in practice. So I will probably only use it as a denoiser.
 

Attachments

  • denoiser12.jpg
    denoiser12.jpg
    326.6 KB · Views: 548
  • denoiser22.jpg
    denoiser22.jpg
    393.9 KB · Views: 550
I guess this layout is more according to Elvee's original vision :D
YESSS!!! Exactly
It is a dienoiser with denoiser stuff in the bottom and dienoiser add-ons on top.
Very clever
However dienoiser seems to be very prone to oscillations and I could not measure any real improvements agains denoiser in practice. So I will probably only use it as a denoiser.
That's strange: I didn't use the dienoiser in reality, but I tested it on a breadboard (which is normally more demanding), and apart from requiring a larger compensation cap, it looked perfectly stable
 
I guess this layout is more according to Elvee's original vision :D
Electrolytics are Nichicon low-profile (100u/35V and 22u/35V).
It is a dienoiser with denoiser stuff in the bottom and dienoiser add-ons on top. However dienoiser seems to be very prone to oscillations and I could not measure any real improvements agains denoiser in practice. So I will probably only use it as a denoiser.



That's so lovely! Amazing work of miniturisation :)

About the oscillation from the Sziklai Pair: Presumably you used a decent C4 value and reasonable ESR in/out caps?
 
I guess this layout is more according to Elvee's original vision :D
Electrolytics are Nichicon low-profile (100u/35V and 22u/35V).
It is a dienoiser with denoiser stuff in the bottom and dienoiser add-ons on top. However dienoiser seems to be very prone to oscillations and I could not measure any real improvements agains denoiser in practice. So I will probably only use it as a denoiser.


Dayum!