Discrete regulators with denoiser

Jump to Latest
#222
PaulSpinks said:
Its just for the supply on 2 OPA1612 gain stages - not the power stage. The 1ET400A DS lists a current requirement of 27mA for the second gain stage, and i'd have thought the buffer gain stage would use a similar amount, so unless i'm missing something we should be well under 100mA.

I'm just not sure whether there is likely much to gain by improving regulation on supply to OPA1612s. I cant find much discussion on how sensitive they are to regulation quality.
Click to expand...
Should work fine but might be overkill for opamps.
 
  • Like
Reactions: mchambin
#224
The TI data sheet gives PSRR figures: 0.1 micro volt per volt ( 140 dB ) typical and 1 microvolt per volt ( 120 dB ).

Further down it gives curves. PSRR+ and PSRR- versus frequecy.
This shows flat 140dB from DC to 10 Hhz then decreasing 20dB per decade.
Hence 120dB at 100 Hz, mains hum in Europe.

This ripple is a sawtooth signal known to have harmonics amplitudes 1/k the fondamental amplitude, where k is the rank of the harmonic.
So, the significant ripple rejection is close to 120dB typical. OPA1612 is pretty good at PSSR.

Unfortunaltely there are other ways for the PSU ripple to sneak in.
A pitfall is the biasing circuitry. Make sure it is designed right.
 
Last edited:
#225
This ripple is a sawtooth signal known to have harmonics amplitudes 1/k the fondamental amplitude, where k is the rank of the harmonic.
Click to expand...
So, the significant ripple rejection is close to 120dB typical. OPA1612 is pretty good at PSSR.
Click to expand...
I am wrong about the 120dB. I overlooked, the rejected harmonics amplitudes do not decay.
The harmonics amplitudes do decay, but the rejection does so.
The result of these two effects is, all harmonics have the same amplitude.
All harmonics in the audio band. There are 200 harmonics up to 20 kHz
The RMS value of the rejected ripple is sqrt (200 ) less, that is 23 dB

So the 100Hz PSU ripple rjection is 97 dB.
 
#226
Spend a little more money and a little more in-to-out voltage drop, upstream of the voltage regulator. Knock down the ripple by 60dB before the voltage regulator ever sees it.

In decibels, PSRR(output) = PSRR(upstream_stage(s)) + PSRR(voltage regulator).
 
#227
Please educate me if I'm wrong here. The diodes hard switching off creates a lot of the harmonics we are trying to block with the PSRR of the upstream + regulator. Then another question, do the active based regulator boards also reduce the harmonics generated? Or am I totally off in left field?
 
#228
The diodes switching, hard or not doesn't create lots of differential mode harmonics, which is what PSRR is about. The waveform is basically a sawtooth, containing even-order harmonics decaying relatively quickly with frequency.
"Dirty" diode switching generates mostly common-mode pollution between the transhormer and mains side and the rectified DC output, and to combat that you need a CM filter
 
  • Like
Reactions: Trileru and RickRay
#229
Mark Johnson said:
Spend a little more money and a little more in-to-out voltage drop, upstream of the voltage regulator. Knock down the ripple by 60dB before the voltage regulator ever sees it.

In decibels, PSRR(output) = PSRR(upstream_stage(s)) + PSRR(voltage regulator).
Click to expand...
My point was not about power supplies PSRR.

My point was about OPA1612 PSRR, a related but different beast.
To hopefully show, there is no need for super duper PSU PSRR to deal with op amps based designs.
 
#230
Hello Trileru thank you for sharing I accidentally found your post because I also happened to be looking for a supply regulator scheme for Wayne's BA2018 scheme. Because I have modified the scheme and layout, I will just change the name according to the version I made, because it is no longer original from Mr Wayne. I need a supply of around 15-25V. You can check the thread about my post here https://www.diyaudio.com/community/...waynes-ba-2018-linestage.423279/#post-7915193
 
#231
Don't think I posted higher Vout values so here's a 30Vout version. Didn't test more than 15Vout so can't say anything about tempco or regulation. I'll try to test 30Vout at some point, and post the results. If 15Vout is too low for you and you want 20-25V I'll post an update.
 

Attachments

  • 30V.png
    30V.png
    90.6 KB · Views: 111
  • Like
Reactions: VIPINSAINI20
#234
Trileru said:
Don't think I posted higher Vout values so here's a 30Vout version. Didn't test more than 15Vout so can't say anything about tempco or regulation. I'll try to test 30Vout at some point, and post the results. If 15Vout is too low for you and you want 20-25V I'll post an update.
Click to expand...
If I use this schematic and create my own PCB layout design, I see AC voltage so I don't need to use a diode bridge, right? So I need to use an AC transformer to connect to the scheme?
And if I use 22V AC voltage from a transformer will the scheme still work properly?
 
#235
For 24Vout you can follow the 30Vout version I posted earlier and maybe drop R6/R18 to 6.8K and adjust Vout with R2/R14.
A 22VAC transformer should work for 24Vout. You need AC to DC bridge, maybe check the AC input PCB designs from post #209 and see how I did it.
If you design your own supply PCB I can't guarantee the performance from the measurements I made.
 
  • Thank You
Reactions: Ujang Ojil
You must log in or register to reply here.
Top Bottom