Sorry i have no schematic , nope only when silence , i`ve already checked all of the psu`s . and they are all ok . I was thinking that the voltage reg my be hissing but checked with other psu and it is the same ...
Or e.g. the UAC2 async controller sends feedback data which are incorrect or unrecognized by the windows UAC2 driver and buffer issues happen on the device.
How did you check they do not occur when music is louder? Are they periodic or random? Do they occur also when no music is being played? When the USB cable is disconnected? How loud are the clicks? Can you record them?
Looks like a clock locking issue then. If the usb stops sending data, the sync is lost. when it starts again it needs to find lock and that gives a click. Keeping sending data (even zeroes) keeps the dac clock in sync.
Take a look at the I2S bus - does it stop? Some controllers will stop I2S data during silence, and most DACs have an auto-mute mode that is engages. That can create an audible click or pop when the clocks resume and the delta-sigma modulators start again. If only DATA becomes zero and the other clock lines stay active, you can sometimes just put an inverter on the DIN line and that usually tricks DACs from ever going to hardware mute.
You could also test by playing another track in the background. Something ultra quiet, the goal is to have some data on the bus, ideally ±1LSB dither. If that stops the clicking, you have an auto-mute problem.
You could also test by playing another track in the background. Something ultra quiet, the goal is to have some data on the bus, ideally ±1LSB dither. If that stops the clicking, you have an auto-mute problem.
If this description is correct:
then IMO there cannot be zeros-only on the I2S data line. It's very unlikely that a silent part of the recording as described in the quote would produce a longer stream of zero samples, for the DAC power management to kick in. But as I say - IF the description is correct. There have been a number of unanswered important questions in this thread.
I`ve tried to capture the best way I can and here is the Output of the DAC playing 500 HZ sine , measured one channel , clicks are in both channels :
Don't know that such measurements are helpful. Scope shots of the click noise in the time-domain view relative to USB board MUTE signal might be helpful.
Those spectrogram lines do suggest a buffer issue, not really related to track volume. Is a reasonable UAC2 feedback value sent? Does the driver respect the feedback value correctly? Both can be determined by analyzing the USB packets e.g. using wireshark, with a bit of effort and subsequent processing.