Super Regulator

I am making a 12v supply for a lumin u1 mini streamer the load is 380ma . Can you suggest a good transformer / rectifier / choke?/ capacitor raw supply for this. Was hoping to achieve a nice sine wave going into the reg as opposed to the saw tooth . Does it pay off to use an r-core or torroidal vs ie core
shottkies? better caps? Is a choke input advisable? thanks
 
I like choke-input filters for steady draw stuff like preamps or class-A amplifiers for a few reasons:
  • They're easier on power transformers since the current draw is more constant than capacitor-input filters', which tend to draw current in narrow spikes. The average current will be the same, but the cap-input filter's higher peaks lead to more I2R losses in transformer windings.
  • Capacitor ripple current after a choke is about an order of magnitude less than if caps are driven right from the rectifier. Once again losses in capacitors are reduced, and their lifetime is extended because of lower ripple.
  • Grounds are never perfect: each ground has some resistance, so the narrow peaky currents in capacitor-input filters require great thick grounding to keep ripple current spikes from contaminating what is supposedly zero volts. For choke-input filters the problem is smaller to the same degree as the reduction in capacitor ripple current. Also the capacitor ripple is essentially sinusoidal instead of the spike train you get with capacitor input filters, so if any hum does exist (which it shouldn't with decent grounding practice) it'll be less offensive since it'll be only at 120 Hz (100 Hz elsewhere), no higher harmonics need apply.

Now nothing comes for free. Chokes have their own issues and it's up to the individual to choose whether they're acceptable:
  • Chokes have losses of their own since they have many, many windings of copper around the core, and low-resistance chokes tend to be non-trivial. You really want to use them at lower than their rated current because otherwise they will run hot, even Hammond parts.
  • As a result choke-input power supplies tend to take up more real estate than capacitor input filters, plus they're heavier to boot.
  • Chokes emit magnetic fields, which can bring up strange problems. For example, I had a persistent low-level hum in my headphone amplifier, which ended up being caused by a ground wire running too close to a filter choke (painstakingly verified by shifting all the other wires and finally in desperation moving the ground lead away from the hum-O-matic).

For me the advantages outweigh the occasional annoyances. Any place I've used a choke has been quieter than an embarrassed church mouse, with absolutely no rectifier buzz.

Plus the darn things look cool. There's nothing like lifting the lid to see a bunch of iron looking back at you.

One more thing: PSUD2 is your friend. It's ostensibly oriented toward high voltage tube stuff but it does just fine on low voltage solid-state things. It's saved my bacon many times and the price is right. Thanks Duncan!
 
More E-choke info here: https://www.diyaudio.com/community/threads/power-supply-mosfet-vs-choke-filter.261117/post-4038197

Elvee shows an interesting gyrator circuit: https://www.diyaudio.com/community/threads/what-is-a-electronic-choke.263220/post-4086812

Of course both of those are capacitor-input power supplies so they still have spiky current draw from the transformer/rectifier assembly compared with honest-to-Cthulhu choke-input supplies, so not quite all of the advantages, more like what tradeoffs one wants to make.
 
Thanks for the suggestions . After some looking it seems that a choke is not needed with a fancy discreet regulator like this . whether there would be any sonic benefit in using a choke I am not sure . My load will be 380ma . Would I be better scaling this regulator up to a 2A version?
 
can anyone recommend some very good sounding caps for my positive reg. . board , I am trying to enhance to sound ...currently I have basic chemicons in it. They are not for audio. Also how critical are the 5 capacitor values. Is 220uf ok instead of 120uf ? I have looked at audionote electrolytics , nichicon UFG, and panasonic os-cons. This will be feeding a streamer. thanks
 
I built this to power a BA2018 line stage.
Simple raw supply with toroid, rectifier, 4400uF filters per side. I never thought the preamp supply could influence the sound as much as this combination.
Much tighter and cleaner. Wow.

I used a 5V LM4040 reference and NE5534 opamps. 2.5Kohm feedback resistor to get to my goal of +/- 18V. Easy. Going to build another to help out my other preamp.

Thank You.
 
@smoovj
just to clarify...
you built a separate raw supply for each channel? so distinct super regulators for each channel too?
and you substituted NE5534s for the standard AD825s?
which resistor is the "feedback" resistor? The chart for setting output voltages specifies an R6/R13 value of ~1.5k for 18V
thanks!
 
No - same supply for both channels.

Yes - subbed the NE5534 for the AD825. You need to be careful here as the 5534 only guarantees stability with gains >3.

The chart is based on a lm329 6.9V reference. 6.9V*(1.5K +1K +100)/1K = 17.94V

I used a LM4040-5V reference in place of lm329. 5V*(2.5K+1K+100)/1K = 18V
 
Variant of the Jung-Didden Superreg with own PCB. Positive 5V version. Unfortunately I did not note where I found that schematic (with this combo of OpAmp, MOSFET etc.). This variant simulated more stable for me than the original, so I went for it. This is of course not to say that the original is less stable, I just had issues simulating it.
Worked well on first try (well actually 2nd... dumb me had reversed the MOSFET). MOSFET is IRF640, voltage ref is LT1004, OpAmp is LT1677.

Thanks @jan.didden for publishing this great regulator. This is now the 2nd after the T-Reg had success with.

1742058445859.png

1742058766490.png