You'll want to put at the very least an LC filter between the supply and your opamp(s).
Try an SLF7045T-471 (470uH) SLF7045T-471MR22-PF TDK | Mouser
in series with the supply rail with a Nichicon HZ UHZ1C102MPM6 Nichicon | Mouser
shunting to 0V on the opamp side of the inductor. Assuming that you're not using more than 16V.
In simulation this combination gives about 80dB attenuation at 40kHz.
Try an SLF7045T-471 (470uH) SLF7045T-471MR22-PF TDK | Mouser
in series with the supply rail with a Nichicon HZ UHZ1C102MPM6 Nichicon | Mouser
shunting to 0V on the opamp side of the inductor. Assuming that you're not using more than 16V.
In simulation this combination gives about 80dB attenuation at 40kHz.
Last edited:
opamps and many discrete amplifiers can have very good PSRR at lower frequencies.
All have poor PSRR at higher frequencies.
The output of a switching supply is riddled with higher frequency interference.
You need something to attenuate the higher frequency interference. A 3pin IC regulator will not do !
All have poor PSRR at higher frequencies.
The output of a switching supply is riddled with higher frequency interference.
You need something to attenuate the higher frequency interference. A 3pin IC regulator will not do !
Thank you for all the inputs, some direction to focus ~ so zero-in action on the right area.
Search in diyaudio, some point out SMPS are usually well-considered design in disturbance reduction; while situation can become more complex when cross influence when supplying multiple devices.
Would appreciate any inputs which can leverage on; what actually was done to migrate and found effective.
Very simple illustration of Andrew's point I wrote-up over a decade ago:
Using 3-pin regulators off-piste: part 2
Note - even a little C-R-C decoupling can help ahead of a 3-pin reg, very much; but you must be careful about where the '0v' ends of those capacitors return to. e
Effectively, they move the noise current from the input to wherever you connect them to - so their 0v termination must be on the 'dirty' 0v side, to return the HF current directly to the raw supply - and not pass across your 'clean' filtered 0v ground routes, anywhere. In other words - it's not enough to pick a preferred approach (LC, CRC, 'gyrator', whatever) - a good, well-considered layout is also an essential part of the solution.
Using 3-pin regulators off-piste: part 2
Note - even a little C-R-C decoupling can help ahead of a 3-pin reg, very much; but you must be careful about where the '0v' ends of those capacitors return to. e
Effectively, they move the noise current from the input to wherever you connect them to - so their 0v termination must be on the 'dirty' 0v side, to return the HF current directly to the raw supply - and not pass across your 'clean' filtered 0v ground routes, anywhere. In other words - it's not enough to pick a preferred approach (LC, CRC, 'gyrator', whatever) - a good, well-considered layout is also an essential part of the solution.
My BD player has SMPS PSU and it is very noisy. I put a number RC snubbers over the SMPS, multiple LCR filters, etc. but found that some ferrite cores wrapping around the output wires which go to the DAC board work wonders. I guess a lot of noise is CM noise so ferrite cores wrapping both V and 0 wires may work better than LCRs that mainly deal with differential noise. Please correct me if I am wrong.
CM noise is only an issue for connections to external units (for example the connection between a preamp and an amp, or between a source and preamp), not for the preamp itself. CM noise between units can be addressed by isolation transformers, slap-on ferrite cores (in effect they're CM chokes) only really help at very high frequencies.
I'd not rely on PSRR alone to reject SMPSU ripple. If you just want to rely on PSRR (and therefore, by extension linear modelling) then since your switching ripple is ultrasonic you can have any amount you like, it won't affect your audio because its above the audio band. But based on subjective observation, ultrasonic ripple does degrade SQ so there must be something else going on - which I take to be intermodulation distortion. The real circuit is somewhat non-linear, which means linear modelling doesn't cut it.
I've drawn a schematic in this post - Anyone tried to build LME49600 Headphone Amplifier Evaluation Board??
How about approach in below link
TIDA-00571 Low Noise High PSRR LDO Reference Design for Powering Hi-Fi Audio Application | TI.com
TIDA-00571 Low Noise High PSRR LDO Reference Design for Powering Hi-Fi Audio Application | TI.com
Is this a solution looking for a problem?
Every response is implicitly saying that the OP's power supply is too noisy and something has to be done to attenuate it or else (implicitly) the sound experience will suffer.
Maybe the OP wouldn't be able to hear any difference. If the post was on a hifi audio forum then its likely 100% of posters would be saying get a better amp (same outcome of a solution looking for a problem).
Just saying
Every response is implicitly saying that the OP's power supply is too noisy and something has to be done to attenuate it or else (implicitly) the sound experience will suffer.
Maybe the OP wouldn't be able to hear any difference. If the post was on a hifi audio forum then its likely 100% of posters would be saying get a better amp (same outcome of a solution looking for a problem).
Just saying
Is this a solution looking for a problem?
Every response is implicitly saying that the OP's power supply is too noisy and something has to be done to attenuate it or else (implicitly) the sound experience will suffer.
Maybe the OP wouldn't be able to hear any difference. If the post was on a hifi audio forum then its likely 100% of posters would be saying get a better amp (same outcome of a solution looking for a problem).
Just saying
isnt 200mV consider noisy for Pre-amp & DAC use?
-- Hear any noise?: Yes, as i can hear little hiss even volume (Potential gain knob) is at Zero, I had relatively quiet environment.
This little hiss tell me, that amplifier is powered-on.
-- 200mV ripple: presumably has some impact to pre-amp, and DAC on SQ.
That sounds like your amplifier is the main contributor to hiss/noise.
If you were inclined to test out the noise/hiss contribution from your preamp, then you need to appreciate that that noise/hiss may be from the preamp electronics, and nothing to do with the powering.
The simplest test for identifying if your preamp power supply is contributing noise/hiss that you can hear is to temporarily use a battery power supply for your preamp.
If you were able discern a difference between battery powering and MeanWell powering your preamp, then you need to somehow eliminate the other plausible causes of that noise/hiss that are unrelated to filtering any high frequency noise (as mostly recommended by the posts in this thread) - one area of concern is ground/HF loops through multiple mains AC connections.
My HP-1 headphone amp uses two 24 V switching bricks + CLC filter + linear regulators to get down to ±18 V. The HP-1 provides world class performance.
My DIFF PRE is powered by two 15 V switching bricks + CLC filter and gets stellar performance as well.
It is perfectly possible to get good audio performance from switching modules chosen with care. The switching frequency of the Mean Well units tends to be around 70 kHz. You can attenuation that with a well-designed CLC filter. Make sure to limit peaking in the filter as that will boost the noise (due to the resulting higher supply impedance) rather than suppress it.
One way to determine whether your preamp or power amp is the dominant noise contributor is to use a shorting plug on the power amp. Short its input. Measure the output noise. Then connect the preamp with the volume turned all the way down (or a shorting plug on the selected input). Measure the output noise. If the noise increases significantly when you attach the preamp, the preamp is dominating (or at least contributes as much as the power amp to the noise level).
Tom
My DIFF PRE is powered by two 15 V switching bricks + CLC filter and gets stellar performance as well.
It is perfectly possible to get good audio performance from switching modules chosen with care. The switching frequency of the Mean Well units tends to be around 70 kHz. You can attenuation that with a well-designed CLC filter. Make sure to limit peaking in the filter as that will boost the noise (due to the resulting higher supply impedance) rather than suppress it.
One way to determine whether your preamp or power amp is the dominant noise contributor is to use a shorting plug on the power amp. Short its input. Measure the output noise. Then connect the preamp with the volume turned all the way down (or a shorting plug on the selected input). Measure the output noise. If the noise increases significantly when you attach the preamp, the preamp is dominating (or at least contributes as much as the power amp to the noise level).
Tom
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.