USB to I2S 384Khz - DSD Converter

[knuisje]

Received today, thanks a lot!

On my selfdesigned ES9016 based dac with 80MHz oscillator and IL715 isolator the 352.8 en 384kHz samplerates produce a lot of noise and odd sounds although the samplerate is correctly identified by the dac.
But all the other samplerates play perfectly without any unlocks, so I'm very satisfied.

[amanero]

To remove the noise you need a very low noise analog supply 1.2V or you need to stop the oversampling in the DAC, 1.5836 Mhz is too high

Via I2C send to the DAC this command address 0x49 register 0x11 value 0x5C

[knuisje]

I immediately tried your suggestion bypassing the OSF filter and I can confirm that 352.8 and 384kHz files now play without any issues.

[glt]

Originally Posted by knuisje

I immediately tried your suggestion bypassing the OSF filter and I can confirm that 352.8 and 384kHz files now play without any issues.

Hmmm, I got to try this. With another interface and Buffalo II 80 MHz, bypassing OSF did not solve the noise problem:
If I turned OSF off, I got the following:

SR<352Khz resulted in fast intermittent sound with the lock-led flashing like crazy. By increasing the dpll bandwidth setting to highest and the dpll bandwidth multiplier to x128 (this is the highest bandwidth setting possible), the lock-led flashing rate slightly decreased as though the DPLL was trying to lock. But there was no lock and no music
For sample rates 352.8K and below, the DPLL was able to lock but only when there was no music content (for example when the player is paused).
SR=384Khz resulted in “normal sound”
The frequency of the DPLL indicated was 64x smaller what it should be for I2S input. So for example, for SR=384Khz, the sample frequency determined by the dpll was 6000 Hz (if you multiply 6000×64 you get 384000 Hz)

[LeonvB]

Received today, thanks!

[knuisje]

Originally Posted by glt

The frequency of the DPLL indicated was 64x smaller what it should be for I2S input. So for example, for SR=384Khz, the sample frequency determined by the dpll was 6000 Hz (if you multiply 6000×64 you get 384000 Hz)

Correct, I now multiply DPLLnum by 64 when OSF Bypass is selected by the user.

[Martin12341]

Received yesterday, thanks. Board is working with Linux Mint 13 without problems.

Just one thing: I would like to use an USB isolator: https://www.olimex.com/Products/Modu...rface/USB-ISO/
This device supports only USB fullspeed but for 44kHz files this should be sufficient. But when I connect the USB isolator the Amanero board is not recognized by the PC. I tested with and without external power supply of the USB isolator, there is no difference. Power supply of the Amanero board is ok, CLK and FSCLK signals are ok, but I2S DATA is permantly high. Any suggestions?

Martin

[qusp]

there is more to signing on with a USB protocol that accepting the current bandwidth required. the device needs to be able to sign on/shake hands as a UAC 2 device

[qusp]

Originally Posted by knuisje

Received today, thanks a lot!

On my selfdesigned ES9016 based dac with 80MHz oscillator and IL715 isolator the 352.8 en 384kHz samplerates produce a lot of noise and odd sounds although the samplerate is correctly identified by the dac.
But all the other samplerates play perfectly without any unlocks, so I'm very satisfied.

no, afaik the 9016 does not actually support the higher rates of its cousins 9012/18.

edit: ahh OK well maybe it does, thats news to me, very interesting! I was planning a design with this dac too (for my subs), but I read even with NOS it didnt support these modes, which would be a problem in my setup, because the other 2 dacs for tweeters and mid-bass (both 9012 based) do

[knuisje]

The datasheet of the 9016 says that the MCLK can be max. 100MHz and for I2S must be min. 192 x Fs.
192 x 384000 = 73.7 MHz so a 80MHz oscillator is good enough.

But the section on OSF bypass says that 8 x Fs is recommended, sourcing data directly into the IIR filter. The maximum samplerate that can be applied is 1.536MHz (8 x 192kHz). This must be the reason samplerates > 192kHz don't work without OSF bypassing.