• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

E1DA Cosmos ADC

ABEB843C-A547-4C10-BBCA-0D5A7291E58D.png
I've checked the solder wirings and turns out that in my rca to xlr converter, the female rca terminal is wired in reverse, so gnd of d30 was connected to hot xlr, etc. etc.

Here's the best that I could get with the Loxjie D30 (e1da right input, i'll try the mono mode next time)
 
I've checked the solder wirings and turns out that in my rca to xlr converter, the female rca terminal is wired in reverse, so gnd of d30 was connected to hot xlr, etc. etc.

It's good you got that sorted out.
 
Here's with Loxjie D30 unbalanced direct to E1DA. I'll try with a different setup later, might be because of wire quality / interference / etc

View attachment 163888
I see here the common noise between USB/ADC GND and DUT GND. Please check how should look the correct cable connection:
Cosmos_ADC2RCA.jpg
2021-11-08_11-09-17.jpg
 
@IVX can you point the location of input protection diodes? I'll check if removing them can lower my high 2nd and 3rd harmonic measurements. Or removing them does not offer any measurement benefits?
 
@IVX can you point the location of input protection diodes? I'll check if removing them can lower my high 2nd and 3rd harmonic measurements. Or removing them does not offer any measurement benefits?
I don't have the ADC but based on the online photo, it should be the one that is on the right of the ESS chip. There should be 8 of them (two diodes for each signal * 2 differential signals per channel * 2 channels).

I would say don't remove them. You'll burn your chip down if you have a higher than expected input signal.
 
jlesterp,
100.00% bad idea for a few reasons simultaneously:
1) I used there 8 diodes in SOD723 case. See attached.
2) I did test that diodes type before use and found they don't affect the distortions.
3) such clumpers never affect the 2nd harmonic.
I calibrate each Cosmos ADC unit with reference sine source <-150db all harmonics, and I have found no one ADC with 2nd harmonic >-140db(3rd >-134db, AVG about -138db). Only one thing may affect the 3rd harmonic level is temperature, I keeping 27C in my room(AC setting) for the calibration process. Today seems getting colder and I gonna warm up the room.
If you have a doubt about my calibration you can try to arrange some AP analyzer for an hour and check this out. However, the precision of AP
is often exaggerated in the consumer's dreams. Take a look at that paper about the sine generator of APx555 investigation https://www.nanovolt.ch/resources/l.../pdf/low_distortion_oscillator_comparison.pdf
2021-11-08_17-26-56.jpg
 
Last edited:
  • Like
Reactions: trl
I don't have the ADC but based on the online photo, it should be the one that is on the right of the ESS chip. There should be 8 of them (two diodes for each signal * 2 differential signals per channel * 2 channels).

I would say don't remove them. You'll burn your chip down if you have a higher than expected input signal.

I have an interface with TPD2E007 for ESD protection and adjustable overvoltage relay protection so I think it can protect the ADC when I measure power amps. Though if I measure other DACs directly, it might get fried (Murphy's Law).

jlesterp,
100.00% bad idea for a few reasons simultaneously:
1) I used there 8 diodes in SOD723 case. See attached.
2) I did test that diodes type before use and found they don't affect the distortions.
3) such clumpers never affect the 2nd harmonic.
I calibrate each Cosmos ADC unit with reference sine source <-150db all harmonics, and I have found no one ADC with 2nd harmonic >-140db(3rd >-134db, AVG about -138db). Only one thing may affect the 3rd harmonic level is temperature, I keeping 27C in my room(AC setting) for the calibration process. Today seems getting colder and I gonna warm up the room.
If you have a doubt about my calibration you can try to arrange some AP analyzer for an hour and check this out. However, the precision of AP
is often exaggerated in the consumer's dreams. Take a look at that paper about the sine generator of APx555 investigation https://www.nanovolt.ch/resources/l.../pdf/low_distortion_oscillator_comparison.pdf
View attachment 164055
Hmm almost the same size as 0402 resistor. I can manage to solder/desolder them but it'll take time.
Looks like more trouble than leaving them alone. I'll follow your advice :)
 
if I'll understand why I need that.
_You_ obviously don't need it since you already have the information. What I've inferred by looking at USB and I2C traffic is that the settings are stored in an EEPROM, and the HID protocol simply provides access to this data, 4 bytes at a time. While the meaning of some of the fields is obvious, others are not. I'm simply curious. If posting the format of this config data is too much trouble for you, or you don't want to reveal it for some reason, then so be it.
 
Because of the discussion about the difficulties of measuring unblanced connections I thought to post a link to Archimago's recent review of the Nobsound ADuM4160 USB Isolator. TL;DR he actually found it to be relatively effective despite its 25$ price. It also has a 24/96 limitation, but it might me useful in some situations.

 
Nor RME (no support for those low speed modes).
 
Good clarifications @IVX and @MC_RME, but it does not have to be used on the ADC. It can help if only used on the DAC. In the review Archimago used it just in front of the Topping D10s DAC and got a 3dB improvement on the THD+N (but still 2,5dB shy of ASR measurements). Measurement system was a single desktop computer (Intel NUC) setup with Topping D10s (with isolator) and RME ADI-2 Pro FS.
 
sseppala, this one USB2 High-speed with 32pF of pass-thru capacitance.
@IVX, I can't follow your meaning, are you saying that Cosmos does not need or does not work with isolator? If you use the isolator only on the DAC which can fall back to USB 1 (and you do not need more than 24/96 output resolution). Doesn't this break groung loop Computer->DAC->ADC->Computer and help with unbalanced measurements. At least Archimago's measurements got improved.

I must admit that I am not an electrical engineer, I have more of a computer science background.
 
That board does USB to/from other interfaces. If what you want is an S/PDIF output from the Cosmos, that will not help.

The various ASRC based boards mentioned above will work if a slight degradation of the signal is tolerable.
I think I finally found a board that can function as I want.
What do you think?
These two products appear to have separate MCLK on the board.


 
I was thinking of using it for both measurements and audio recording. Will test it with friends on several devices as soon as I have time.
Regarding the audio recording, I thought about another way :
- using the Cosmos ADC to record via it's USB
- using the I2S to clock signal only and let Cosmos ADC be the Master and set the DAC as slave, sync with the clock coming from Cosmos ADC

Maybe @pkane or @IVX can confirm if something like this can work : aliexpress.com/item/4001143681130.html
H9b7ff1667a7644f39539fc8cc9b80fbey.jpg
Did you find a solution?
I found a board with a separate MCLK. If you need it, please refer to it.


 
Back
Top Bottom