Hello
I have the following problem.
When I'm listening to music, I hear the noise of the speakers - at very low levels of the sound card with an amplified to a maximum amplifier.
When I measure the signal with an oscilloscope and connect a capacitor - this signal disappears .
Which one is better to use in the output of my sound card - boucherot cell or low pass . how to calculate its value.
This image shows an audio signal - when the sound card is at 0% volume. And this signal makes noise into the speakers.
I have the following problem.
When I'm listening to music, I hear the noise of the speakers - at very low levels of the sound card with an amplified to a maximum amplifier.
When I measure the signal with an oscilloscope and connect a capacitor - this signal disappears .
Which one is better to use in the output of my sound card - boucherot cell or low pass . how to calculate its value.
This image shows an audio signal - when the sound card is at 0% volume. And this signal makes noise into the speakers.
Attachments
I believe the correct translation is you hear a non-musical sound in the speakers, not of the speakers.
Don't know what a boucherot cell is. However, a one or two pole low pass filter should help since this noise is much above the audio band. The period on the scope is 10 usec, so it is a 100 khz signal.
So a 50 khz inflection point low pass filter is your desire.
I'd try a zobel type coil in series with the output hot, followed by a capacitor to output ground. Output of this to your powered speaker.
A zobel type coil is 11 to 16 turns of wire around a AA battery or a china marker. For the capacitor I'd try a .01 uf film parallel a 33 or 47 pf ceramic cap. The reason for the ceramic, the coil may pick up RF interference. Those figures aren't precise or calculated, any old salvage film cap over 1000 pf might work.
If you put the whole filter in a grounded steel box, it would reduce the chance of picking up RF. Ground of the box would be the safety ground of the house outlet, or some convenient water or gas pipe.
For calculating, see wikipedia complex impedance. Z impedance of a capacitor is ~1/(j*omega*C) where omega is frequency*2*pi and C is in farads.
Z impedance of an inductor (coil) is ~j*omega*L in henries. A 14 turn zobel coil is about 5 microhenry.
Series impedance is vector sum of two part impedances. A scalar approximation is sqrt(z1^2+z2^2)
The capacitor parallel to ground has impedance reciprocal to one in series.
See why I didn't calculate it? Just take your junk box of caps & wire and go to it. No junk box of caps? pick up an old radio out on the curb on garbage day, with leaded parts (pre 2000) and disassemble it. If you disassemble a TV, short out the electrolytic caps and the tube anode without touching them with your fingers. Use a shorting probe, Alligator clip (to device ground) series wire series 600v meter probe (from dead meter). See high voltage safety sticky thread under tube amps forum for picture.
Don't know what a boucherot cell is. However, a one or two pole low pass filter should help since this noise is much above the audio band. The period on the scope is 10 usec, so it is a 100 khz signal.
So a 50 khz inflection point low pass filter is your desire.
I'd try a zobel type coil in series with the output hot, followed by a capacitor to output ground. Output of this to your powered speaker.
A zobel type coil is 11 to 16 turns of wire around a AA battery or a china marker. For the capacitor I'd try a .01 uf film parallel a 33 or 47 pf ceramic cap. The reason for the ceramic, the coil may pick up RF interference. Those figures aren't precise or calculated, any old salvage film cap over 1000 pf might work.
If you put the whole filter in a grounded steel box, it would reduce the chance of picking up RF. Ground of the box would be the safety ground of the house outlet, or some convenient water or gas pipe.
For calculating, see wikipedia complex impedance. Z impedance of a capacitor is ~1/(j*omega*C) where omega is frequency*2*pi and C is in farads.
Z impedance of an inductor (coil) is ~j*omega*L in henries. A 14 turn zobel coil is about 5 microhenry.
Series impedance is vector sum of two part impedances. A scalar approximation is sqrt(z1^2+z2^2)
The capacitor parallel to ground has impedance reciprocal to one in series.
See why I didn't calculate it? Just take your junk box of caps & wire and go to it. No junk box of caps? pick up an old radio out on the curb on garbage day, with leaded parts (pre 2000) and disassemble it. If you disassemble a TV, short out the electrolytic caps and the tube anode without touching them with your fingers. Use a shorting probe, Alligator clip (to device ground) series wire series 600v meter probe (from dead meter). See high voltage safety sticky thread under tube amps forum for picture.
Last edited:
Thank you very much for your posts.
In a few days I will have time to calculate and assemble a filter against the problem.
These 100kHz have them at a 0% sound level. It may be from the sound card that is Creative X-Fi Xtreme Music , and perhaps from the power supply that is Fortron Bluestorm 500w.
In a few days I will have time to calculate and assemble a filter against the problem.
These 100kHz have them at a 0% sound level. It may be from the sound card that is Creative X-Fi Xtreme Music , and perhaps from the power supply that is Fortron Bluestorm 500w.
No, plain ol' 4556A + 4558 if memory serves. I bet the noise seen on the scope is an artifact of the ground loop inherent to the measurement setup, making some power supply noise visible (unbalanced connection between two Class I devices, the old story).
If I got that right, the OP is saying that there is noise to be heard when the amplifier is cranked all the way up - to which I'm saying yes of course. A typical integrated amplifier (which I'll assume is being used) has about 14-20 dB more gain than you really need with a soundcard like this, as input sensitivity for full power output tends to be in the 150-200 mV range (40-46.5 dB of voltage gain).
The XtremeMusic had like what, 106 dB(A) of SNR? Referred to 2 Vrms, that's like 10 µV of output noise. Plus 46 dB, that's 2 mV or about 13 dB more than the inherent noise floor of an average integrated amp (add another 10 dB for a good one). I'm not at all surprised if that's plainly audible.
In short, you should not need to crank up the volume beyond about 12 o'clock (or ~20 dB below maximum). If that's fine, (non-)problem solved. Classic case of unrealistic expectations. Sometimes it does pay to do the math.
If I got that right, the OP is saying that there is noise to be heard when the amplifier is cranked all the way up - to which I'm saying yes of course. A typical integrated amplifier (which I'll assume is being used) has about 14-20 dB more gain than you really need with a soundcard like this, as input sensitivity for full power output tends to be in the 150-200 mV range (40-46.5 dB of voltage gain).
The XtremeMusic had like what, 106 dB(A) of SNR? Referred to 2 Vrms, that's like 10 µV of output noise. Plus 46 dB, that's 2 mV or about 13 dB more than the inherent noise floor of an average integrated amp (add another 10 dB for a good one). I'm not at all surprised if that's plainly audible.
In short, you should not need to crank up the volume beyond about 12 o'clock (or ~20 dB below maximum). If that's fine, (non-)problem solved. Classic case of unrealistic expectations. Sometimes it does pay to do the math.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.