[ti5002000]

b0bb
Did you find ainy significant improvements when switching to the Belleson regulator?

 

[MartinWT]

It depends of your preference, for something smoother try a polystyrene cap, if you prefer a brighter and sharper presentation try a Corning CGW Glass Cap.

Thanks. I meant to ask, what value are they?

I prefer smoother, do you have any high quality polystyrene caps in mind that would fit in that location?

 

[b0bb]

Thanks. I meant to ask, what value are they?

I prefer smoother, do you have any high quality polystyrene caps in mind that would fit in that location?

Try Ebay
100pF on Gen 1, 200pF on Gen2, connect 2 100pF caps in parallel.

https://www.ebay.com/itm/293630525981

 

[b0bb]

b0bb
Did you find ainy significant improvements when switching to the Belleson regulator?

Incremental improvement, clearer upper midrange and soundstage.
You will not see significant improvement until the input and output caps are upgraded.

 

[Xoverman]

3 minor mods to SU-6.

Made SU-6 much smoother and organic, an unexpected surprise.
https://www.head-fi.org/threads/singxer-su-6.902272/page-30#post-16233262

Yes, I had the same results
Most of it comes from the polymer capacitors.

 

[Xoverman]

So it's only when you use DSD?
You are using the volume of the 004 ?
There is no preamp. The 004 is directly connected to the power amp. ?
Noise on both channels when volume is very low?
Noise is input independent !
Noise on all outputs ( R/L , RCA / XLR ) ?
Next try the independent +/- legs off the XLR agenst GND. Is the noise one all 4 amp's ( PINS) or just on one.

 

[b0bb]

Rebuilt LKS Amanero interface

1. Swapped 22/24MHz clocks for 45/49MHz, reduced jitter by 50% (3dB). CCHD957 family lowest phase noise around 45-49MHz, pickup addditional 2dB of noise reduction.
2. 12MHz 10ppm MEMS XO for USB microcontroller, to improve the USB receive performance
3. Replace regulators : ADP151 (PSRR: 70dB, 200mA 9 μVRMS noise) with TI TPS7A20xx (PSRR: 95 dB, 300mA, 7 μVRMS noise )
4. Replace 4.7uF ceramic regulator input filter cap with 47uF Tantalum Polymer for better noise filtering
5. Tantalum Nitride resistors for I2S Data, Bit Clock, Frame Clock/DSD data lines
6. Tantalum Nitride coupling resistors for XOs
7. Tantalum Nitride USB signal termination resistors
8. Silicon RF cap (10pF,1%) for USB PLL filter

Resistors are the blue rectangles

Board prior to mod

 

[ESL57]

CVHD-950 is a Voltage Controlled XO, any noise pickup on that pin strong enough to raise the voltage causes a shift in the output frequency.

Should not be used for the 004.

CCHD-950X is out of stock at the major suppliers.
You could try the SIT-5357 but it is 2-2.5x the cost and much more difficult to solder onto the board.

This product code is the correct one from SIT5357 clock :
https://www.digikey.es/product-deta...473-SIT5357AE-FQ-33N0-100-000000F-ND/12033037
If this is it, it is also out of stock in the usual sites
Also I will change the 74HC244 of origin for the suggested ones and some other capacitor in the regulators.

 

[ESL57]

So it's only when you use DSD?
You are using the volume of the 004 ?
There is no preamp. The 004 is directly connected to the power amp. ?
Noise on both channels when volume is very low?
Noise is input independent !
Noise on all outputs ( R/L , RCA / XLR ) ?
Next try the independent +/- legs off the XLR agenst GND. Is the noise one all 4 amp's ( PINS) or just on one.

So it's only when you use DSD?..........Yes, only with DSD, but with PCM they are not audible but with an oscilloscope they can be seen very very weak
You are using the volume of the 004 ?........ Not always at 0db or bypass mode, I have a preamp
There is no preamp. The 004 is directly connected to the power amp. ?........... I check in diferents preamps
Noise on both channels when volume is very low? .......Random noises on the 2 channels, they never repeat in time if you play them different times comparing them. Noises spoil the listening in these files recorded with little level, they are audible at 0db or byppas mode (maximun)
Noise is input independent ! ....Yes
Noise on all outputs ( R/L , RCA / XLR ) ?..... Yes, They are already in the resistors i / v
Next try the independent +/- legs off the XLR agenst GND. Is the noise one all 4 amp's ( PINS) or just on one....... I would have to see it again, I think I remember that I put the oscilloscope on the 4x 68.1ohm resistors and the noises were already there.
I tried the same resistors as your 3w high thermal stability and less ohm and with others of more ohm, I only managed to decrease and increase the gain respectively, with 75ohm the noises are a little more masked.

 

[b0bb]

This product code is the correct one from SIT5357 clock :
https://www.digikey.es/product-deta...473-SIT5357AE-FQ-33N0-100-000000F-ND/12033037
If this is it, it is also out of stock in the usual sites
Also I will change the 74HC244 of origin for the suggested ones and some other capacitor in the regulators.

Get it from Mouser
https://www.mouser.es/ProductDetail/SiTime/SiT5357AI-FQ-33N0-100000000X?qs=gTYE2QTfZfSK84/PcXFFEA==

The 74ALVC244 from Nexperia/NXP is the one in the photo.
It has special signal conditioning at the inputs (Schmitt Triggers), this reduces jitter and hopefully and potential DPLL lock issues.
Neither TI or ON mention this in their datasheets, so get the specific one from Nexperia
https://www.mouser.es/ProductDetail/Nexperia/74ALVC244D118?qs=P62ublwmbi8nefVp98M2BQ==

There are a lot of diffrences between the logic families (HC, AC, LVC, LVT, ALVC etc) do not substitute the ALVC.

 

[piaseczek]

@ESL57 I listened to the problem audiofile you attached.

There was a burst of noise followed by a click, this pattern repeats.
It might be the DPLL not being able to maintain a stable lock.

Check your I2S cables and connector on the 004 to see if anything got broken.

Also check the I2S receiver chip and the 74xxx244 line drivers

The chip handling I2S and USB data is the left one in the picture.
(I am in the process of swapping these out for the 74ALVC244)

b0bb,
Buffers in my 004 look different.
Should it be replaced?

 

[b0bb]

So it's only when you use DSD?..........Yes, only with DSD, but with PCM they are not audible but with an oscilloscope they can be seen very very weak
You are using the volume of the 004 ?........ Not always at 0db or bypass mode, I have a preamp
There is no preamp. The 004 is directly connected to the power amp. ?........... I check in diferents preamps
Noise on both channels when volume is very low? .......Random noises on the 2 channels, they never repeat in time if you play them different times comparing them. Noises spoil the listening in these files recorded with little level, they are audible at 0db or byppas mode (maximun)
Noise is input independent ! ....Yes
Noise on all outputs ( R/L , RCA / XLR ) ?..... Yes, They are already in the resistors i / v
Next try the independent +/- legs off the XLR agenst GND. Is the noise one all 4 amp's ( PINS) or just on one....... I would have to see it again, I think I remember that I put the oscilloscope on the 4x 68.1ohm resistors and the noises were already there.
I tried the same resistors as your 3w high thermal stability and less ohm and with others of more ohm, I only managed to decrease and increase the gain respectively, with 75ohm the noises are a little more masked.

What operating system software and player are you using to send DSD to the DAC?
004 is quite fussy about the quality of the DSD sent to it.

Have you tried HQPlayer ?
Time limited demo is free, available for Linux/Windows/Mac.

Try it with Amanero, no driver needed for Linux/MAC, ASIO driver needed for Windows.

 

[b0bb]

b0bb,
Buffers in my 004 look different.
Should it be replaced?

Post a picture, I think yours was 74AC244D, lower current output version of the 74HC244D.
74ALVC has higher current output.

74ALVC244 has better noise tolerance and better drive capability.
The propagation delay is 3X better than the 74AC244

74AC244 is used so that cheaper parts like the ceramic caps below can be used to save on manufacturing cost.

I think you should eliminate the software player as possible cause before modifying the hardware.

If you have the 74AC244, and want to use the 74ALVC244 the ceramic cap in the red squares have to be replaced with Polymer caps.
I use the 820uF Nichicon FP-RNL

 

[piaseczek]

AC244 ON CPAE0805 - Google knows nothing about it...

 

[b0bb]

AC244 ON CPAE0805 - Google knows nothing about it...

AC244 is the partnumber, the other might a batch or manufacturing tracking identifier