I don't have a preamp, I don't have anything hooked up. The noise is with nothing connected to the ncore input.
I've figured out the loud pop is because of the mytek, but the static that I have on my tweeter has nothing to do with the mytek. The ncores make the white noise with absolutely nothing hooked up. With DAC/PRE or without, my speakers still exhibit the same static/white noise.
robbbby...
I am confused: I thought you said that with the inputs shorted, there is no "bad" sound at the tweeters? If this is the case, then there is no problem with your amps. When then input is shorted, it stops RF pickup through the input (from airborne sources).
I also am a little bit confused by your use of the words "static" and "white noise". Static, I would describe as a scratchy sounding noise, not smooth in nature, with a definite rough texture. White noise, on the other hand, would be a very smooth, even sounding noise, with no texture: by definition I believe white noise is comprised of an even distribution of noise all across all frequencies, with no one frequency at a higher level than another, it should be smooth and even.
In my experience, the self noise of the nCores on their own is smooth and even textured very soft hiss, what I would describe as white noise at an extremely low level. A static like sound, would indicate a possible problem if you had that sound with the input shorted. With the inputs open, they are available to pick up airborne sources of RF, and I would not necessarily conclude that there is a problem, even if you hear a rough noise floor.
I am confused: I thought you said that with the inputs shorted, there is no "bad" sound at the tweeters? If this is the case, then there is no problem with your amps. When then input is shorted, it stops RF pickup through the input (from airborne sources).
I also am a little bit confused by your use of the words "static" and "white noise". Static, I would describe as a scratchy sounding noise, not smooth in nature, with a definite rough texture. White noise, on the other hand, would be a very smooth, even sounding noise, with no texture: by definition I believe white noise is comprised of an even distribution of noise all across all frequencies, with no one frequency at a higher level than another, it should be smooth and even.
In my experience, the self noise of the nCores on their own is smooth and even textured very soft hiss, what I would describe as white noise at an extremely low level. A static like sound, would indicate a possible problem if you had that sound with the input shorted. With the inputs open, they are available to pick up airborne sources of RF, and I would not necessarily conclude that there is a problem, even if you hear a rough noise floor.
The way I understand it is that with inputs shorted, the hiss/crackle/whatever goes away, but with open inputs (no preamp or DAC) it is there.
I agree totally open inputs will pick up noise, the output impedance of the preamp or DAC should change that. Might be worth testing with a resistor of a few kohm across the input connector pins.
Yes, it is a hiss, full of treble. A hiss with all octaves equal in energy is pink noise: the flat white noise filtered by 3dB/oct on all the band.
[edit] Typo corrected.
Last edited:
I guess it depends on if you are talking amplitude/voltage dB's, or power/loudness dB's...
They're the same dB's. 3dB is double power or 1.414x voltage which is the same thing
White noise has a constant power density per unit bandwidth (per Hz). Pink noise has a constant power density per LOG of unit bandwidth, e.g. per octave or per decade).
So, white noise has as much energy between 0-200Hz as between 10000Hz-10200Hz. Imagine we cut up the audio range:
Bass: 20-200Hz (one decade, 180Hz)
Mid: 200Hz-2000Hz (one decade, 1800Hz)
Treble: 2000Hz-20000Hz (one decade, 18000Hz)
Pink noise will put exactly equal amounts of power into the three ranges. White noise will put 9 times more power in the mids as in the bass, and yet another 9 times in the treble as in the mids.
How does that work out in "powerdeebees"? Well, 10*log(9)=9.5dB. 9.5dB more in the midrange than in the tweeter. 9.5dB more in the midrange than in the woofer.
How does that work out in "voltagedeebees"? Well, 9 times the power is 3 times the voltage right? So, 20*log(3)=9.5dB. 9.5dB more in the midrange than in the tweeter. 9.5dB more in the midrange than in the woofer. No difference, see?
Okay. So how does white noise sound? Well, considering that per the above white noise contains practically no bass it sounds thin, hissy and tweetery.
How do you make white noise? The simplest method is a resistor. Resistors are sources of white noise. Look up Johnson Noise for more info. Noise voltage goes up with the square root of R. Suppose you have an amplifier with a 100k resistor strapped across it. Measured over the audio range the amp sees 5.6uVrms noise. Suppose you then connect a low resistance source, say 10 ohms. Total resistance, pretty much 10 ohms. The amp now sees only 56nV.
Without source: tweetery noise. With source: quiet.
How do you make pink noise? Find a waterfall. Failing that, use a white noise source and something we call a "pinking filter". That's a 3dB/octave filter. Remember: every octave has twice the bandwidth as the one before that. So white noise puts twice the power in that higher octave. Or 1.414x the voltage. 3dB either way. So the filter needs to be something special. A 1st order filter is too steep. How to make a half order filter? You can't, but with alternating real poles and zeros you can get as close as you want.
White noise has a constant power density per unit bandwidth (per Hz). Pink noise has a constant power density per LOG of unit bandwidth, e.g. per octave or per decade).
So, white noise has as much energy between 0-200Hz as between 10000Hz-10200Hz. Imagine we cut up the audio range:
Bass: 20-200Hz (one decade, 180Hz)
Mid: 200Hz-2000Hz (one decade, 1800Hz)
Treble: 2000Hz-20000Hz (one decade, 18000Hz)
Pink noise will put exactly equal amounts of power into the three ranges. White noise will put 9 times more power in the mids as in the bass, and yet another 9 times in the treble as in the mids.
How does that work out in "powerdeebees"? Well, 10*log(9)=9.5dB. 9.5dB more in the midrange than in the tweeter. 9.5dB more in the midrange than in the woofer.
How does that work out in "voltagedeebees"? Well, 9 times the power is 3 times the voltage right? So, 20*log(3)=9.5dB. 9.5dB more in the midrange than in the tweeter. 9.5dB more in the midrange than in the woofer. No difference, see?
Okay. So how does white noise sound? Well, considering that per the above white noise contains practically no bass it sounds thin, hissy and tweetery.
How do you make white noise? The simplest method is a resistor. Resistors are sources of white noise. Look up Johnson Noise for more info. Noise voltage goes up with the square root of R. Suppose you have an amplifier with a 100k resistor strapped across it. Measured over the audio range the amp sees 5.6uVrms noise. Suppose you then connect a low resistance source, say 10 ohms. Total resistance, pretty much 10 ohms. The amp now sees only 56nV.
Without source: tweetery noise. With source: quiet.
How do you make pink noise? Find a waterfall. Failing that, use a white noise source and something we call a "pinking filter". That's a 3dB/octave filter. Remember: every octave has twice the bandwidth as the one before that. So white noise puts twice the power in that higher octave. Or 1.414x the voltage. 3dB either way. So the filter needs to be something special. A 1st order filter is too steep. How to make a half order filter? You can't, but with alternating real poles and zeros you can get as close as you want.
You are right, of course - my brain fart, caused by lack of enough morning coffee.
20/20 Mr Bruno (as usual and we can feel your jubilation when it is about poles.
)That is something which always has surprised-me, this difficulty to generate a pink noise with those ugly filters.
This energy repartition (1/ƒ noise ) is pretty natural in many physical or biological systems, and our senses works near that way...
When pink noise appears in electronic devices (Flicker noise ) it is always where you don't want because the corner frequency is set by Mr Murphy.
nb: Zeners are pretty noisy, and can be used to generate white noise more easily than with resistances, despite the amp noise itself is not so boring when amplifying resistance noise. ;-)
Mr Murphy allow us to use zeners in power supplies and other reference voltage stages, where we want... No noise.
Last edited:
Sorry, i would describe it as white noise, it's just a even sound.
And yes with inputs shorted no sound, but open (nothing connected, no dac, no preamp), I get the white noise.
So this is totally normal with nothing plugged in? The input should take care of that noise? So if it still makes white noise with my combination dac/pre plugged in, the noise is because of my dac?
I would make one more test - I would try a 1Kohm resistor between the - and + inputs on the input XLR connector to see if the noise goes away.
Oh, and one more thing - could you try disconnecting the input cable from the nCore (connector J9) to check if the noise changes?
if you have a high output impedance DAC and a less than optimal layout I'd say you should do something about it.
single-ended input? improper (not twisted, unshielded, shield not properly connected) input cables running too close to the inductor or high current traces? what DAC? layout (pics)?
also, how bad is it? how sensitive are the speakers, how audible is it (distance)?
PS: I took a look at the audiocircle NCORE builds thread and I simply can't understand why some go to extreme lengths with casework and such but at the same time use "audiophile" output wires left untwisted. somehow I was surprised to see that balanced inputs were used at all.
Last edited:
Last edited:
that would explain it, but still... properly twisting the wires would take 1 minute and not connecting them on opposing sides takes just as much effort as doing it.
anyway, I'm trying to get my head around something. there are enough reports about the NCORE being substantially better than the UCDs in the bass dept. this correlates with my personal experience, the UCDs sounding bass shy compared to a good class A amp. sometimes (with double bass for instance) I feel that a high-pass filter has been applied and lower tones are missing, thus making the instrument sound not exactly real.
what I'm unable to do is correlate this with the measured performance of said amps. anyone have any idea? even speculation would be better than nothing.
oh, and to robbbby: I searched your posts and looks like you have the Mytek 192 DAC and (according to you) it has a high (100-200 ohms) output impedance. that would explain your noise issues.
Last edited:
- Status
- Not open for further replies.