CM6631A + ES9023 cheap board - modification...?

[chermann]

Hi

i bought this board to have a cheap USB/SPDIf converter...more or less it lay around...the sound is not bad for that money !! (thx to abraxilato for that hint)

2017 CM6631A Digitale schnittstelle 32/24Bit 192 Karat Soundkarte USB zu I2S/SPDIF Coaxial Ausgang Unterstutzung Connect Decoder Upgrade DAC in 2017 CM6631A Digitale schnittstelle 32/24Bit 192 Karat Soundkarte USB zu I2S/SPDIF Coaxial Ausgang Unterstutzung Connect Decoder Upgrade DAC aus Verstarker auf AliExpress.com | Alibaba Group

before i start with mods on other boards i want to learn on this cheaper board
yes i go throug some other posts and foud some advices for the ES9023 chip....but i found nothing about my board.

Es9023:
its on the board with 2 caps to GND (220/100 marked) and according to the datsheet (pin 7+ 8 Aoutl+AoutR) they are not needed so i move them. i had a 2nd board but after some cap changes i never works again....so i re cheked all caps compared to origin bought caps. so all caps are good and measure with my LCR some a origin WIMA. so the quality is fine. only these 2 caps at are strange (at 10khz i go 220 but ESR is"OL" and the LCR meet jumped to RP mode....) these caps are at the buttom yellow pianted...neverthe less you do not need it

pin 6 Vreg:
here i found on the board no setup resistor which recommended in the DS -R8 should set 130k (for 3,6V - 2Vrms output) or 220k (for 3,3V - 1,9Vrms output)...just a caps is there so what do here?

PIN 15 Mute: this is set to +VCC with the pull up of 100k according to the Datasheet...ok

the voltage for all chips on board comes from the AM1117 3,3V
shold i change this to AP2114 or LT1963 or ADP151 ???

what about the internal charge pump? its the internal "generator" from e.g. 3,3V to an + / - voltage for the internal opamp in the ES9023 chip. is there a modification to do?

the driver in Win10 is working.if i go into the setup/advanced i can see the C- media driver and if i want to change /update the driver i got the positive feedback that my driver is the latest...so fine Tidal is use the "USB 2.0 HD audio device"


if somebody is interest to help me you are very welcome

chris

 

[chermann]

I2S mode...

Hi

I dont check it on the board but i guess the DIF Pin4 on the ES9023 should be set for the communication to the CM6631A via I2S.
so the PIN 4 - DIF should be set according to the datsheet -to GND:
Input to select Left Justified or I2S data
Driving the DIF pin low will set the DAC in I2S mode while
driving the pin high will set the DAC in LJ mode.
correct?
what about this modes...can somebody give me some explanations please?

MCLK
i checked the PIN 13 MCLK (Masterklock) at the Es9023 and its connected with the CM6631A.so it should be ok. the reson wha i am askin is that on guy in the net removed this MCLK(13) completly to force the ES9023 in async mode......hhmmm...what t..

thank you in advance
chris

 

[chermann]

Hi to all..

after reading a lot off pages my summary is:

headache


to be honest i read all pages of that:
https://www.diyaudio.com/forums/digital-line-level/197486-measurements-es9023-dac.html

https://www.diyaudio.com/forums/digital-line-level/331959-cheap-chi-fi-dac-buy-30-usd.html

Hifimediy ES9023+TE7022 24/96 USB Dac

ES9023 DAC PCB

Bragi, an Asynchronous ES9023 DAC with isolated I2S Input and Alps RK27 Volume pot

this are partly "old" posts 2012-2013 so mybe some ideas are not updated. but i jsut concentrate not on the reciever chip now i look at the ES9023 first.

so my main conclcusio is:
power up the ES9023 with 3,6V (jean paul really stick on this in all posts )
if you want to keep the board as it is just change the 3,3V regulator (its for a ll chips) to a better= low noise version
which on i cant say

use the Vreg pin 6 with a resistor the valus is recommnded with 130k but lot of guys use 100kR and are happy with this

split if possible -its a decision of keeping the case ore build a "bigger" dac - the regulator of all the chips

not so clear for me:
about caps changing and
Cn+CP for the internal voltage pump its different-so no idea
caps at the output (PIN 7+8)..its writiten in the DS = 4700pF but a ot of guys use nothing


mybe i am years behind ?- lot of actual cheaper boards are available with ES9018

bye chris
....hopefully somebody will jump in and help me here

 

[chermann]

...first mod done..Vreg resistor...

Hi

After some compare sessions with my Tooping D50 (its not needed to power on my Gustard A20H) i made the first really compare session yesterday evening.
i use Tidal master recordings to compare. i use 5 titles and connect 4-5 time for a fair comparison.
The D50 is powered / no mods (just -4dB volume settup) with an power pack (poweradd 10000mAh) and its new but is used it for 40hours i guess.the CM6631A board is running about the same time so its ok.
the mods are:
no caps at the outputs (post 1) and i add a non smd restiors under pcb the cap at Vreg (see pic)..3,3V power for the chip its recommended to use 220kR....but i just have a 180,1kR measured...so i try it.


Sound check: Windows setup is 24/192k for both devices
Without Vreg R:
the sound of the cheap board is nice not tonal mistakes, room is deep but i miss more darkness, soundstage ok. no harshnees, no blub bass.....fine for that price. btw... if you use the alu case for that board ... the mistake of L+R channel is correct shown on the alu front.... funny...
D50 :
if i switch on and checked the setting in Tidal /Windows an start playing its immediatly noticable the the darkness and the kind how instruments/voices can have more room for living and a better swing in/swing out....ok its about x10 the price...now its clear that the separation and th focus of voice/instrumnts is missing for the CM6631/9023 board.

With Vreg R=180k:
to be honest i do not expect a change.......but.....yes the seperation and focus is better....first i really dont belive it and i switched additionally 3 times to approve that phenomenum...meanwhile it was past 10:00PM but i can recognize this changes.


nice sounding board i can say...


chris

 

[chermann]

...ohh...i forgot to write...
i second board is on the way. this provide me to verify the modifications in the future better between some steps

 

[chermann]

...mod 2 - regulator change...

hi

here is the plan for the voltage regulator update - its instead of the AM1117-3,3V. this better regulator - i focused on AP2114j...this is a overview ...

red critical
yellow could fit
green fine



its the first shoot.....so.....any inputs are welcome....

chris

 

[chermann]

AMS1117 regulator values

Hi

i have to wait for the second board for better A/B comparerision....
i want to show you the case. (pic 1+2) ....here you see the correct out L+R...

meanwhile i checked the ams reg. values which are on the board. --> pic3

Cin =470µF/25V
Cout=470µF/25V + red wima with 0,015µF(foil cap)

according to the datasheet its not ok. i found for the input no value so i will keep it.
Cout:
here i found just 22µF tantal or some post 10µ ESR about 30mR...so look at pic 4. so my next step will be to keep this regulator but change the values what is recommended......or i have into the box.....incl. measurements.

if i write nonsense ...please let me know..


chris

 

[chermann]

Adum4160 win setup and sound check

Hi


Waiting for the other board i start with the topic USB isolator
i bought a ADUM4160 about 18 Euro at amazon
DollaTek USB Isolator 1500V Isolator ADUM4160 USB-zu-USB-Isolatormodul Audio-Rauschunterdrucker Industrieller Isolatorschutz: Amazon.de: Elektro-Grossgerate


the Sound check with the CM6631A board and my Topping D50 DAC is successfully. as i read the max audio format with this adum is 24Bit/96kHz...fine i can l live with that.

the adum is directly connected after the laptop and powered with a different USB port as external power supply...therefore you have to set the switch.
pic 1



first i start with topping D50 and i take 1 setting and reboot to have win recognize the adum+D50.
i tried a month ago a ifi audio i usb3.0 nano (220 euro) but had no big approvements ! this adum is really good and i can really recommend this box
sound gets sorted, darker, bass controlled, more "right" sound


CM6631A board
first.......... every connect/disconnect force a notable pop in my chain -speakers
the max audio format after i tried several time with win setting is 24/48kHz...its not possible to get more... pic 2
any comments or hint to solve that is very wellcome



nevertheless the sound improvement is really good....the difference is the same as at the D50...the small DAC get more air and 1 step deeper sound stage


so the low cost DAC grow from 35euro +18euro(adum)....


chris

 

[phase]

Do the clocks on that board have any local decoupling?
I didn’t see anything, but thought maybe there was something on the underside of the board.

 

[chermann]

Do the clocks on that board have any local decoupling?
I didn’t see anything, but thought maybe there was something on the underside of the board.

Hi phase


you are the first here.....thank you


no.... under the pcb there is nothing to see.....see pic

yesterday i retry several timse but without adum4160 the cm6631A board is able to set to 24or32bit/192khz and if i switch the adum in the chain tidal is requesting an error and its just "allowed" to set to 24b/48k


but sound check is still with adum better.
yesterday evening i tried 10 times with settings 24 or 32 bit /192k without adum and play cm6631A directly...its better then 24/48k. 32/192k sound a little bit strange in my opinion..24/192k is more darker the 24/48k but with adum4160 its better sound stage, clear highs, the character of an instrument get very clear.....more holographic sound.


e.g. natural instruments...very good recording...deep soundstage
bueno vista social club /title "dos gardenias" + "pueblo nuevo"


chris

 

[chermann]

hi phase


i forgot to write
the link i posted..if you click on the pics and zoom in you have good resolution to see the smd components


thanks
chris

 

[jean-paul]

I think phase meant these. They are lacking on the pictures so there is a chance for improvement as XO's should be decoupled.

 

[jean-paul]

BTW you are right in using a better regulator as AM1117 is not very good. You could keep the AM1117 for powering the XO's and a separate better 3.6V version for the ES9023 together with a 130 kOhm for Vref. Please use the output filter caps as you assumed wrongly that they are not needed. Beware of assuming things. Which kind of output cap for the regulator (and what value) depends on which regulator you will use. Never assume and change and then use the ears to decide as it does not work like this when decoupling regulators. This method works when coupling analog stuff.

As far as I can tell there is only one regulator on the PCB powering all things. Using a separate regulator for the ES9023 will be a significant update.

 

[chermann]

Oh...i miss interpret the adum with the cm6631A board...thanks jean paul...you are very welcome here.


yes i read a lot of things in many pages ....so wrong way is for a noob normal...
i am really happy that you have time to help me for this modding.


chris

 

[jean-paul]

Thanks. Please note that the XO power supply pins are opposite to eachother in a diagonal. You probably can use SMD caps straight from the + pin to the GND plane if you scratch away a bit of the blue coating. I would use 0.1 µF 0603 (if that is correct, please check size).

The 1117 is famous for noise which is (depending on manufacturer) around a whopping 100 µV. I would replace it anyway also when it only feeds the XO's.

 

[chermann]

...try to look at the dac noise...

Hi



i use my scope and a differential probe and set the measurement point at the amps input rca ( its a amp board..easier to measure, its the same amp as i use for listening now).

no music amp baord is muted. dac on, amp on
pic 1 win setting 24bit/48khz , no adum4160
pic 2 win setting 24bit/48khz , with adum4160
so the difference is here but not as big...



but with win setting 24/192k..its better ?!?

pic 3 win setting 24bit/192khz , no adum4160


are these measurements ok?
chris

 

[jean-paul]

They do not look OK to me. First do the suggestions and remeasure.

 

[chermann]

...measurments at the AMS1117...with usb isolator

i try to measure the effect of the ADUM4160 at the ams1117 input and output


pic 1 with adum4160 input pin
pic 2 without adum4160 input pin
pic 3 with adum4160 output pin
pic 4 without adum4160 output pin


not really clear for me..slightly better in the rms values...but nothing "special"
so the ams1117 is noise...nobody can help him



chris

 

[chermann]

They do not look OK to me. First do the suggestions and remeasure.

solder is on.....i go in my lab on solder the 220pF (no 4700pF foil available at home)


can you explain what you can see here...



thx
chris

 

[jean-paul]

This particular board exists in various versions. Does yours have the beads in the power supply lines? You definitely have power supply issues so a good first would be replacing the 1117 and give the ES9023 a separate reg.

Please use the datasheet values 4.7 nF for output filtering.