I would like to ask whether a capacitance multiplier circuit can suppress interference, more precisely, ringing, from a switching power supply. The latter is a laptop charger power supply which supplies power to a laptop which I use for a USB oscilloscope. Interference signals are showing superposed on whatever signal I happen to examine and view. These are in the form of decaying ringing. Their effect also prevents whatever signals are viewed to lock into view.
You might wat to look here.
https://www.diyaudio.com/community/...all-warts-preamps-hpa-korg-nutube-etc.354213/
Not exactly an answer to your question but maybe an answer to your problem.
https://www.diyaudio.com/community/...all-warts-preamps-hpa-korg-nutube-etc.354213/
Not exactly an answer to your question but maybe an answer to your problem.
The interference in these cases is mostly caused by common-mode perturbations, against which a differential mode filter or a cap-multiplier is ineffective.
You need a common-mode filter, based mostly on a number of CM chokes, each dealing with a part of the spectrum. A DM filter can do no harm though, and could even improve matters slightly
You need a common-mode filter, based mostly on a number of CM chokes, each dealing with a part of the spectrum. A DM filter can do no harm though, and could even improve matters slightly
Like LV saz, switch-mode noise is high frequency and ground based, not across the power supply. Run the power supply wires, ie including ground through a ferrite core, several turns. Be sure your switch-mode line input filter is grounded. Assume there is a ground current that needs blocking. You have a RF voltage on all the power supply outputs with respect to the ambient ~ground, not across those outputs. You cannot short the noise to your analog ground, you have to block it with inductance. All wires from the power supply must have a "common mode" inductance, ie they are wound together.
That.
Capacitance multipliers, or plain old extra raw capacitance, deals with ripple, but not HF switching trouble which can get into your amp in sneaky ways.
A gret part of the problem is using a dirty power supply NOT designed for Audio.
Capacitance multipliers, or plain old extra raw capacitance, deals with ripple, but not HF switching trouble which can get into your amp in sneaky ways.
A gret part of the problem is using a dirty power supply NOT designed for Audio.
What could also be happening is that switching hash is picked up by the o'scope probe ground lead. Try twirling it up around the probe body to minimize the loop area. If that doesn't work, you may have to run your laptop on battery power when you're using the scope.
Tom
Tom
Thanks for replying. Since, the interfering signals are not between the power supply terminals, a "differential" inductance with two coils seems to be the solution. The question is: what value of inductance should I use? I have some ferrite cores which I can use, but the constant of inductance, that is, the inductance with one turn, is very low.
Typical values of commercial CMC's are generally comprised between 2 x 1.5mH and 2 X 47mH, and the value is correlated to the frequencies targeted.
Ideally, you would want a very large value to cover the widest possible frequency range, but in practice large inductors have a large parasitic capacitance, degrading their HF/VHF performance.
A workaround is to use a multisegmented coil former, like this one:
But there is always a limit, and to properly cover the whole spectrum you need to cascade a number of graduated coils: for example just a simple ferrite tube to begin with, followed by a 2 x 22µH, a 2 X 220µH, 2 X 2.2mH, 2 X 22mH.
It is desirable to add some form of damping too, as uncontrolled resonances can have weird effects, like differential to common mode conversion
Ideally, you would want a very large value to cover the widest possible frequency range, but in practice large inductors have a large parasitic capacitance, degrading their HF/VHF performance.
A workaround is to use a multisegmented coil former, like this one:
But there is always a limit, and to properly cover the whole spectrum you need to cascade a number of graduated coils: for example just a simple ferrite tube to begin with, followed by a 2 x 22µH, a 2 X 220µH, 2 X 2.2mH, 2 X 22mH.
It is desirable to add some form of damping too, as uncontrolled resonances can have weird effects, like differential to common mode conversion