I heard, some audio cards are using so called 'jack sensing' feature, to determine, what was plugged in - microphone or headphones, etc... Maybe this does produces such a problem? Is this HF signal presented at the card output all the time during PC startup? Is this noise present during playback of some track at low volume level?
Very interesting problem
Very interesting problem
No. This problem is due to the nature of sigma-delta converters. All sigma-delta converters has more or less out of band noise, while in band noise is suppressed by a so-called noise shaping. Out of band noise is suppressed by analog filter, but since in this case the pass-band is wide, it can't be filtered out as good as in a low sampling rate DAC. Also the oversampling ratio is lower, what makes the situation worse.
You can measure up to 300Khz with Audigy 2 Value sampling at 192Khz? How does that get done?
I thought so also, but now...
Hi,
when measure output noise what is parameters (samples/sec)?
I know Audigy 2,flat-108dB up to 46KHz.
noise have relation with samples/S (non linear) because board use multiplier clock divisor . plus dac not professional.
Unfortunately I can't adjust the sampling rate (nor the filter roll-off, what makes me more sad), but I played a 4...90 kHz sweep before measurement, and it passed, so sampling rate must have been 192 kHz. But the noise is the same (more or less) when I use winamp, with 44,1 kHz MP3. Would be good to be able to adjust DAC manually!
You can measure up to 300Khz with Audigy 2 Value sampling at 192Khz? How does that get done?
No, as I said I measured it with oscilloscope.
Hi,
I have curiosity,sorry, why can't adjust parameters of audigy?
this is installed on pc? (normally have software for set samples/s etc).
mybe use at external custom ?
---------------
yes,for class D is necessary filter 24dB for remove carrier(at -30dB out of band).
measure with sound card is not attendible,specially on fft .
stranous frequency peak on fft,are "immaginary" result of AD-DA clock process in relation a some real peak frequency at input (class D have).
for thd at 1Khz is good (for referement measure).
I have curiosity,sorry, why can't adjust parameters of audigy?
this is installed on pc? (normally have software for set samples/s etc).
mybe use at external custom ?
---------------
yes,for class D is necessary filter 24dB for remove carrier(at -30dB out of band).
measure with sound card is not attendible,specially on fft .
stranous frequency peak on fft,are "immaginary" result of AD-DA clock process in relation a some real peak frequency at input (class D have).
for thd at 1Khz is good (for referement measure).
Of course I measured only the soundcard yet.
If I connect this noisy signal to a Class D amp, the noise is sampled by the amplifier and can be transponed to audible band, this is why I mentioned Class D amp.
I guess it's hard to understand what problem you really have, and what tests you are doing. Normally, it's best to look at the noise spectrum graphically to make a better judgement on what is happening. I would assume that either you have an input directly feeding the output internally, creating some sort of loop; or it's just the computer ground plane is not clean.
Done. I didn't take a photo, but the noise is between 60 and 300 kHz.
No external input is enabled for playback. Only wave out is enabled.
It would have make an audible noise, but this is not. And 3 of 3 independent PC with the same, incredibly strong GND problem is very unprobable.
Don't guessing, the source of the noise is obvious already, only the quantity (the way to make it less) is questionable.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.