• 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

audiofun

Active Member
Joined
Jul 4, 2021
Messages
242
Likes
243
me too, if anyone can explain it would be nice.

the explanation is clear enough.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
audiofun, The explanation is about how but I wondered why sources should be synced? For example, if Fs=48kHz for all ADCs connected but they are not synced, what is the problem?
Why the audio SW can't simply record a few files simultaneously if the time of their start is known? Next time, when you simultaneously play these files-tracks via a mixer they will be synced with microseconds(1/48000s=20us) of the mismatch vs real guitar/drums/vocal performance. Who cares about 20us, or even 2.5us at 384kHz?
 

phofman

Senior Member
Joined
Apr 13, 2021
Messages
489
Likes
319
Each 48kHz is different. A good example is the recent log of capture/playback rate correction in CamillaDSP at https://audiosciencereview.com/forum/index.php?threads/rpi4-camilladsp-tutorial.29656/post-1225165 . The average rate ratio between incoming SPDIF stream (i.e. clocked by the SPDIF transmitter) and outgoing stream via USB adaptive (i.e. clocked by the USB controller) in this particular setup is 100.02%:

2022-06-22 09:57:48.670793 DEBUG [src/audiodevice.rs:614] Current buffer level: 784.7, corrected capture rate: 100.0250%

That makes 10 samples difference every second at 48kHz. A buffer setup at 50ms for a reasonable non-interactive latency will over/underflow in 2400/10 = 240 secs = 4 minutes. Much faster if some interactive work is required and the buffer must go down to 5-10 msecs max.

Of course for measurements the generator and analyzer can be independent and do not need synchronization. Yet e.g. generating 1kHz in the setup above means there would be 1.0002kHz at the analyzer.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,013
Location
Berlin, Germany
Of course for measurements the generator and analyzer can be independent and do not need synchronization.
That sure is true for many of the standard audio measurements, but sometimes a hard sync between ADC and DAC clock (integer ratio) is beneficial and sometimes even a prerequisite (when time-domain averaging has to be used).
The lack of syncing option is one of the reasons I didn't go for a Cosmos ADC as of yet. I am aware that Ivan has provided a sync signal on an pin to derive a SPDIF carrier signal from with some glue hardware, which then can be used to sync an ADC -- but I'd prefer an out-of-the box solution.... maybe sometime in the future a sync output or input (with adjustable multiplier, even) will come...
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Each 48kHz is different. A good example is the recent log of capture/playback rate correction in CamillaDSP at https://audiosciencereview.com/forum/index.php?threads/rpi4-camilladsp-tutorial.29656/post-1225165 . The average rate ratio between incoming SPDIF stream (i.e. clocked by the SPDIF transmitter) and outgoing stream via USB adaptive (i.e. clocked by the USB controller) in this particular setup is 100.02%:

2022-06-22 09:57:48.670793 DEBUG [src/audiodevice.rs:614] Current buffer level: 784.7, corrected capture rate: 100.0250%

That makes 10 samples difference every second at 48kHz. A buffer setup at 50ms for a reasonable non-interactive latency will over/underflow in 2400/10 = 240 secs = 4 minutes. Much faster if some interactive work is required and the buffer must go down to 5-10 msecs max.

Of course for measurements the generator and analyzer can be independent and do not need synchronization. Yet e.g. generating 1kHz in the setup above means there would be 1.0002kHz at the analyzer.
Cosmos ADC uses ES9822Pro as the master clock source and 10ppm oscillators, how high the time error we can expect in that case?
BTW, if anyone needs the ASIO for Cosmos AADC in Windows, StaticV3 found some strange but tested way to install 2 Comtrue drivers one by one and finally got the native ASIO works.
So the voodoo is as follows:
a) Install the latest Comtrue driver from here https://www.comtrue-inc.com/index.php/download2/summary/4-ct7601/12-windows-7-10-driver
b) Let Windows restart and install another one from there https://drive.google.com/file/d/11LDzaULXWbtP1fyH6j7sp2S2SOF5ve3x/view?usp=sharing
Probably needed to restart again but in my case, it doesn't.
REW works with Comtrue ASIO and no error was reported.
 
Last edited:

audiofun

Active Member
Joined
Jul 4, 2021
Messages
242
Likes
243
yeah, I calculated something too and decided to ask phofman.
your adc's clock may have very low drift. but user's dac's may not. So you need to use resampling when the dac cannot sync clock with your adc.

also bear in mind that 10ppm is the drift number. user may also use a low drift number clock but the clock rate may not be the same as yours.
 
  • Like
Reactions: rsy

Atanasi

Addicted to Fun and Learning
Forum Donor
Joined
Jan 8, 2019
Messages
713
Likes
792
Ideally different audio interfaces would not have phase differences either. This is important if multiple microphone inputs are mixed together. I don't know if resampling can even fix phase differences.
 

LutinBleu

Member
Joined
May 31, 2022
Messages
6
Likes
3
Location
Canada
A newbie question : I am considering buying the E1DA Cosmos ADC to connect a microphone to my ADI-2 DAC FS. What cable should I use knowing that the USB on the DAC is already taken by my MacBook?
 

staticV3

Master Contributor
Joined
Aug 29, 2019
Messages
7,539
Likes
12,005
A newbie question : I am considering buying the E1DA Cosmos ADC to connect a microphone to my ADI-2 DAC FS. What cable should I use knowing that the USB on the DAC is already taken by my MacBook?
What's the signal chain exactly?
Mic->Cosmos ADC->MacBook->RME->Speakers?

Edit: You can't turn your RME into an Audio Interface by plugging the Cosmos ADC directly into it. It doesn't work like that.
You can turn your MacBook into a machine for both Audio recording and playback by plugging both an ADC and a DAC into it via USB.
 
Last edited:

LutinBleu

Member
Joined
May 31, 2022
Messages
6
Likes
3
Location
Canada
What's the signal chain exactly?
Mic->Cosmos ADC->MacBook->RME->Speakers?
I don't know, I was thinking Mic > Cosmos ADC > RME > Genelec 8030 and Macbook > RME, if that makes sense. I haven't thought about connecting the Cosmos directly to the Macbook.
 
Last edited:

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
people keep asking me about APU PCBA adjustment, that's actually simple but I prepared some help about that:
2022-07-22_11-18-29.jpg


and the simulation file for the Microcap 12(it is a free simulator).
 

Attachments

  • Cosmos_APU_Notch_trimpots.zip
    203.3 KB · Views: 72

2pi

Member
Joined
Jul 31, 2022
Messages
5
Likes
7
Location
South Germany
Trying to get ready for measurements with the ADC, APU and Topping HS-01 as required (using REW).

Question about the DIP switches:

Are the switches configuring simple voltage dividers ? If so, could I get the resulting resistance values for each sensitivity setting, please ?

I assume the 1.7V setting has no divider in the way and the signal goes straight to the ADC inputs, right ?

The AUX input seems to be subject to another fix attenuator. So, same question here.

What is the max. load the APU can drive ?

Any news about the DAC ? That would be higher on my shopping list than an autoscaler :)

Oh, and maybe an inverse RIAA for the ADC could also be a nice toy.
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,682
Likes
4,220
Location
Liège, Belgium
people keep asking me about APU PCBA adjustment, that's actually simple but I prepared some help about that:
View attachment 219623

and the simulation file for the Microcap 12(it is a free simulator).
Thanks
It's actually pretty easy to tune it.
I've tuned mine to 997Hz -30dB...

(AES-17 defines frequency for most measurements to be 997Hz.
20th harmonic then falls below 20kHz)
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
2pi, I think you did miss that https://e1dashz.wixsite.com/index/cosmos-adc
2021-07-24_19-12-54.jpg

About Cosmos DAC and Cosmos Scaler, the last one is ready but I have doubts if I can sell >100pcs the same as APU, so I think I'll not even start the production so far.
Cosmos DAC is a complex machine due to the super-fine LPF set(I use the proto to test Cosmos APU), so I afraid to spend 1-1.5 years to design that and sell 100pcs again, sorry, perhaps Cosmos DAC will never be released too. At least for now, I have such a feeling, let see later if I'll change my mind about it. People asking me for portable DACs, so I'll better be focused on that trivial products ;)
 
Top Bottom