E1DA Cosmos ADC

[charliecola]

One problem solved and another pops up (it's worth it though - this ADC sounds amazing) - thank you Ivan.

When I don't install the driver everything works with regards to the UAC2 volume slider in windows - I set to zero and get stereo input fine - it's remembered when restarting computer. The problem with no driver is that I can't get 24 or 32 bit through ASIO4ALL which is a necessity.

When the driver is installed the volume slider in windows becomes stuck at zero - it simply doesn't move - I have to use the tweak app to change to stereo but when the computer is restarted the Cosmos is back in mono mode - is there any way to fix this and have normal UAC2 control with the driver installed so I don't have to use the Tweak app every time I start the computer?

 

[staticV3]

@charliecola Please open the Device Manager, go to "Sound, video and game controllers", right-click the Cosmos -> Properties -> Driver.
Here's what you should see:


Driver provider: Comtrue
Driver version: 6.0.1.4031

 

[charliecola]

@charliecola Please open the Device Manager, go to "Sound, video and game controllers", right-click the Cosmos -> Properties -> Driver.
Here's what you should see:

View attachment 193104
Driver provider: Comtrue
Driver version: 6.0.1.4031

Hi - that is exactly what I see - when this driver is installed I cannot enter stereo mode through windows because the volume slider is stuck at zero (despite it being at zero every time I restart the computer the Cosmos is in mono mode) - with that driver installed I must use the Tweak app to enter stereo mode every time I start the computer. I need to use that driver though so that I can get 32 bit through ASIO4ALL.

Thank you for the help.

 

[Grooved]

Grooved, huh, that looks like you've killed I2S input/s of the CT7601 or output/s of ES9822. Probably ESD. Pls take a look with a scope is there normal I2S signals?

Thanks @IVX , I will check that.
I tried to really take care of each mod I made, and which is strange is that after the last one, it was still working via USB. The only problem I got where that I never got a perfect signal from I2S (was near at a moment with just a bit of distorsion on the digital signal from the digital board.
I don't understand why it was working two days ago, even one hour after the last mod, left it, and two days later it doesn't work.
Can ESD act later like that?
I'm now thinking about too many plug/unplug (from the Cosmos I2S) of the connector I made. I should have made it fix or left it plug on the Cosmos ADC, and plugging/unplugging only on the digital board

 

[staticV3]

@charliecola
Please try uninstalling the currently installed driver and installing this one.
With that driver installed, you should see the Level control in the sound control panel switch from dB values to simple numbers, indicating that Windows has recognized the UAC2 capabilities of the device and has switched from software volume control (dB) to hardware volume control (numbers).
Indeed, that's what I'm seeing on both my Ryzen laptop and my intel PC.

 

[charliecola]

@charliecola
Please try uninstalling the currently installed driver and installing this one.
With that driver installed, you should see the Level control in the sound control panel switch from dB values to simple numbers, indicating that Windows has recognized the UAC2 capabilities of the device and has switched from software volume control (dB) to hardware volume control (numbers).
Indeed, that's what I'm seeing on both my Ryzen laptop and my intel PC.

That is the driver that has this problem for me - it happens on the laptop too, without the driver (using Windows driver) the volume slider operates as expected and the Cosmos memorises it's in stereo mode (or a least windows send out the signal to put it in stereo mode) - it's only once the driver is installed that the Windows volume slider is stuck at 0... as pictured. I would happily run without the driver but then I don't get 32 bit throughput with the audio application.

 

[IVX]

Grooved, not necessary ESD but GND to GND voltage/currents may kill the inputs. You have to connect GNDs of boards first and other signals next to eliminate the chance to kill inputs.

 

[Grooved]

Grooved, not necessary ESD but GND to GND voltage/currents may kill the inputs. You have to connect GNDs of boards first and other signals next to eliminate the chance to kill inputs.

Hi @IVX
it's what I did first, and I tried to do anything with care, but after changing plug, GND was not alone anymore and GND and I2S signals were plugged at the same time.
I got same measurements than before on I2S pins, so it was not on ES9822Pro output, but between I2S pins and ComTrue chip.
I said "was" because after not using it for some days, I worked on it again trying to get signal from the digital output board even if there was nothing via USB, and after removing the resistor on MCLK path, I did another try on USB, and it work again !
So, in the end, I'm not sure of what happened.

Now, I still have not found how to get a good signal.
Sometimes, the recording interface linked via SPDIF to the digital board locks instantly, sometimes, it's after some minutes... and I get various results:
- ComTrue ASIO and WASAPI look to give same results
- with digital output board, if I use internal clock of the recording interface, I get added distorsion but the file has the same length
- still with digital output board, if I use SPDIF clock (locked on the SPDIF signal), no distorsion, but file is 4s shorter on a 2min original file (and pitched)
if I manually change the length, I ended up with a file that has 41dB difference with original in Deltawave, instead of 58dB if I record via Cosmos ADC USB (ComTrue ASIO or WASAPI)

I may also try to create a Word clock signal from LRCK, and locked the recording interface with it while recording the SPDIF signal. Maybe it would work better

P.S. in case it came from that, what do you think would be the best connecting/powering order?

 

[IVX]

the order is only one: connect/disconnect first, turn On next. I have no idea why Comtrue "understands" I2S from ES9822 correctly but your SPDIF board doesn't. Probably SPDIF board has the serial audio setting for nonI2S format(do exist a few variants left, right, I2S).

 

[mansr]

the order is only one: connect/disconnect first, turn On next. I have no idea why Comtrue "understands" I2S from ES9822 correctly but your SPDIF board doesn't. Probably SPDIF board has the serial audio setting for nonI2S format(do exist a few variants left, right, I2S).

Those formats are similar enough that you'd still get a seemingly valid signal, though the sample values would of course be garbled.

 

[Grooved]

the order is only one: connect/disconnect first, turn On next. I have no idea why Comtrue "understands" I2S from ES9822 correctly but your SPDIF board doesn't. Probably SPDIF board has the serial audio setting for nonI2S format(do exist a few variants left, right, I2S).

Thanks @IVX
Turning On last is what I do each time for sure, just wasn't sure about if it would be better to turn On the digital output board first and then Cosmos ADC, or the opposite

Those formats are similar enough that you'd still get a seemingly valid signal, though the sample values would of course be garbled.

Any idea on why the record via SPDIF with the interface locked on it (when this works as it's not always the case) being faster (and pitched)?
I'm searching how it can have ended up like that

 

[Music1969]

A few more THD dB's squeezed out of the DAC sine generator (ADI-2 Pro FS --> Cosmos ADC). These all appear to be limited by the DAC generating the test


32 multitone:

View attachment 160400

Hi @pkane @KSTR can you please run this same 44.1kHz multitone with your RME , but running the Cosmos at 384kHz and showing full bandwidth.

I want to see if the anomolie I see at 88.2kHz is my RME or the Cosmos anti-aliasing filter. Or something else like Windows.

When I get back home i will share mine too

(not an audible concern - just didn't expect to see that strangeness at 88.2kHz with ADC running at 384kHz)

 

[IVX]

Guys, for about a month I wasn't able to do my work and APU/Scaler designs were delayed as well. The reason is the same that makes many of you wake up and every morning check the news about Ukraine where innocent people suffering under putin's aggression.
Anyhow, I am close to the final version of both these products and want to ask your opinion regarding the APU notch performance. At the moment I have the choice to make APU 1kHz notch with -145db/-160db or -150db/-150db 2nd/3rd harmonics resolution. I did ask that in our discord and as I see people prefer the -150/-150 version.
This is about the opamp choice, OPA2156 makes -145db/-160db, Chinese opamp RT6863/SGM8261 -150db/-150db(less OLG and worse linearity but better input stage symmetry vs OPA2156) and a lot cheaper. Of course, I tried OPA1612 as well, some of them gave -148db/-155db but 2nd harmonic is noticeable varies from chip to chip, and I don't like the idea to test at such nanovolts levels each unit. 10kHz notch -130/-130 at least both variants. Please, let me know what do you think about that choice.

The green plot is a passive notch with corrections -50db for the 2nd and -54db for 3rd harmonic i.e. -158/-161db.
The yellow plot is APU result, the correction is -30db for all harmonics(its notch is active) i.e. -150.53/-151db.

And APU with OPA1612.

 

[pkane]

Hi @pkane @KSTR can you please run this same 44.1kHz multitone with your RME , but running the Cosmos at 384kHz and showing full bandwidth.

I want to see if the anomolie I see at 88.2kHz is my RME or the Cosmos anti-aliasing filter. Or something else like Windows.

When I get back home i will share mine too

(not an audible concern - just didn't expect to see that strangeness at 88.2kHz with ADC running at 384kHz)

As I use a Mac, I'm limited on the sampling rate with REW at 192k.

This is Multitone playing a 32-tone test signal into ADI-2 Pro FS @44.1kHz sharp filter, measured by Cosmos ADC at 176.4k

 

[pkane]

Guys, for about a month I wasn't able to do my work and APU/Scaler designs were delayed as well. The reason is the same that makes many of you wake up and every morning check the news about Ukraine where innocent people suffering under putin's aggression.
Anyhow, I am close to the final version of both these products and want to ask your opinion regarding the APU notch performance. At the moment I have the choice to make APU 1kHz notch with -145db/-160db or -150db/-150db 2nd/3rd harmonics resolution. I did ask that in our discord and as I see people prefer the -150/-150 version.
This is about the opamp choice, OPA2156 makes -145db/-160db, Chinese opamp RT6863/SGM8261 -150db/-150db(less OLG and worse linearity but better input stage symmetry vs OPA2156) and a lot cheaper. Of course, I tried OPA1612 as well, some of them gave -148db/-155db but 2nd harmonic is noticeable varies from chip to chip, and I don't like the idea to test at such nanovolts levels each unit. 10kHz notch -130/-130 at least both variants. Please, let me know what do you think about that choice.
View attachment 196162
The green plot is a passive notch with corrections -50db for the 2nd and -54db for 3rd harmonic i.e. -158/-161db.
The yellow plot is APU result, the correction is -30db for all harmonics(its notch is active) i.e. -150.53/-151db.

And APU with OPA1612.
View attachment 196177

Hi Ivan, I prefer the -145/-160.

 

[Music1969]

This is Multitone playing a 32-tone test signal into ADI-2 Pro FS @44.1kHz sharp filter, measured by Cosmos ADC at 176.4k

Can you run the Cosmos at 192k? Hard to see the anomaly at 88.2k there. Need higher bandwidth (if possible).

 

[pkane]

Can you run the Cosmos at 192k? Hard to see the anomaly at 88.2k there. Need higher bandwidth (if possible).

Looked exactly the same @192k.

 

[Music1969]

Looked exactly the same @192k.

Noted but what do you see a little after 88.2? Can't be exactly the same because current plot cuts at exactly 88.2k ?

 

[pkane]

Noted but what do you see a little after 88.2? Can't be exactly the same because current plot cuts at exactly 88.2k ?

Oh, right. Forgot you were asking about 88.2k. Probably won't help you much, even at 192k, since the ADC filter cuts at around that frequency:

 

[pkane]

Noted but what do you see a little after 88.2? Can't be exactly the same because current plot cuts at exactly 88.2k ?

Hooked up a PC to measure at 384k. What is it that you see at 88k? Different than below?