Ciao everybody,
I am playing around a ES9018 based DAC and I am finding problems when i play traces over 44.1KHz sampling rate.
At 44.1KHz everything is fine. Using any input, SPDIF or USB.
At 88.2 and 96KHz the signal is locked but I have in the background a kind of "light rain against a window" sound. Not very loud but clearly present.
If I play 176,4 KHz or over traces the signal is still locked but I have a kind of white noise over the whole sound program. The sound is not destroied but added to the orginal sound there is a nuisance "white" noise.
Cannot be the cables because the trouble is the same with USB, BNC, RCA and TOSLINK.
So it is the DAC board. It works fine with the input sockets installed on the mother board.
But I can't use in the original way because I am trying to use an existing cabinet.
How to connect the input sockets to the DAC existing sockets holes? Now they are connected by a simple wire 3 cm long, and the sockets are installed on the rear panel. A picture is attached. Should be the "wiring" I use? I know it is impossible to do worse but I am trying to use an existing cabinet.
Where start to investigate? Has somebody experimented a similar behaviour?
Thanks.
I am playing around a ES9018 based DAC and I am finding problems when i play traces over 44.1KHz sampling rate.
At 44.1KHz everything is fine. Using any input, SPDIF or USB.
At 88.2 and 96KHz the signal is locked but I have in the background a kind of "light rain against a window" sound. Not very loud but clearly present.
If I play 176,4 KHz or over traces the signal is still locked but I have a kind of white noise over the whole sound program. The sound is not destroied but added to the orginal sound there is a nuisance "white" noise.
Cannot be the cables because the trouble is the same with USB, BNC, RCA and TOSLINK.
So it is the DAC board. It works fine with the input sockets installed on the mother board.
But I can't use in the original way because I am trying to use an existing cabinet.
How to connect the input sockets to the DAC existing sockets holes? Now they are connected by a simple wire 3 cm long, and the sockets are installed on the rear panel. A picture is attached. Should be the "wiring" I use? I know it is impossible to do worse but I am trying to use an existing cabinet.
Where start to investigate? Has somebody experimented a similar behaviour?
Thanks.
Attachments
Ciao canever!
Since no one has responded yet, I guess I'll mention an observation about my own Sabre device, in case it is somehow helpful.
On mine, I currently have a 40MHz MCLK. It looks like you have a 100MHz MCLK, but perhaps my findings will still be helpful. As I understand, ESS specifies that I need at least an 80MHz MCLK for 192KHz support. Nonetheless, I remember in the past I accidentally ran 176.4 to it a few times, and it did lock.
However, when locked @ 176.4KHz, it was not exactly well behaved. If there was a lot of > 40KHz noise in the source audio (such as broadband noise from delta-sigma modulation), I get sound that resembles what you describe. That is, it would not sound distorted exactly, but would sound like white noise is laid on top. If I filtered at ~30KHz or so, or there just wasn't much high-frequency noise to begin with, it would sound (more or less) OK. Otherwise, noisy.
So, with this in mind, let's try something out with your case. Try playback @ 88.2, 96, and 176.4, but apply a (steep) FFT filter beforehand using a DSP program (e.g. Audacity, Adobe Audition, or something like that). Perhaps you will find something similar to what I described, and this will help narrow down the problem.
Since no one has responded yet, I guess I'll mention an observation about my own Sabre device, in case it is somehow helpful.
On mine, I currently have a 40MHz MCLK. It looks like you have a 100MHz MCLK, but perhaps my findings will still be helpful. As I understand, ESS specifies that I need at least an 80MHz MCLK for 192KHz support. Nonetheless, I remember in the past I accidentally ran 176.4 to it a few times, and it did lock.
However, when locked @ 176.4KHz, it was not exactly well behaved. If there was a lot of > 40KHz noise in the source audio (such as broadband noise from delta-sigma modulation), I get sound that resembles what you describe. That is, it would not sound distorted exactly, but would sound like white noise is laid on top. If I filtered at ~30KHz or so, or there just wasn't much high-frequency noise to begin with, it would sound (more or less) OK. Otherwise, noisy.
So, with this in mind, let's try something out with your case. Try playback @ 88.2, 96, and 176.4, but apply a (steep) FFT filter beforehand using a DSP program (e.g. Audacity, Adobe Audition, or something like that). Perhaps you will find something similar to what I described, and this will help narrow down the problem.
Last edited:
Thanks for your reply.
I am running with 100MHz MCLK, the maximum possible and XO is left floating.
Anyway I understand should be a noise summed somehwere to the original signal. What kind of filter you will apply? Low pass? and the frequency? Should be something makes the MCLK dirty?
Unfortunately my oscilloscope is rated 200MHz and it is impossible to investigate on a 100MHz MCLK.
I am running with 100MHz MCLK, the maximum possible and XO is left floating.
Anyway I understand should be a noise summed somehwere to the original signal. What kind of filter you will apply? Low pass? and the frequency? Should be something makes the MCLK dirty?
Unfortunately my oscilloscope is rated 200MHz and it is impossible to investigate on a 100MHz MCLK.
Try lowpass filter at 20KHz, then maybe 25, 30, 40, and then no lowpass. Do this in DSP (that is, apply lowpass using a computer program for example). See if the noise changes.
- Status
- Not open for further replies.