Cosmos APU a notch+LNA $70 to outperform APx555b for $30,000

[KSTR]

Has anyone else seen this with ESS DAC chips?

I'm not sure I do understand the problem.
But I can state the Cosmos ADC with selected brickwall filter (Linear Phase Apodizing) is perfect in not allowing any aliases to come through (100dB attenuation).
Whatever is in the spectrum it is coming from the DUT alone.

 

[tombo56]

Cosmos ADC - ES9068AS

 

[IVX]

Just back to update the new avatar

 

[tombo56]

Ghostsbusters
THDbusters.

 

[TNT]

What about THD?

//

 

[IVX]

no THD, almost zero.

 

[TNT]

Aha - you avatar means dead to THD rather than THD is a bad idea for measurement...?

//

 

[IVX]

right, depends on the particular point of view. I, personally, feel that the PS quality affects the SQ much more. However, a bad power quality always brings extra distortions, hence, the THD test is correlated with SQ even indirectly.

 

[IVX]

It seems the first 9039S is burned by shorting its output to the GND. Guys, please remember:

9039S has balanced outputs, Cosmos ADC as well, don't make "adapters" like this RCA pls!

 

[Markw4]

Okay. Here is what a google search gives:
https://www.google.com/search?q="A&...mAIAoAIAmAMAiAYBkgcAoAct&sclient=gws-wiz-serp


Which one would you like people to use?

 

[IVX]

this one

 

[IVX]

but the problem isn't about the pinout. All R-/R+/L+/L- four outputs are active opamps outputs that's why the shorting isn't allowed.
BTW, yesterday I selected one of the best THD 9039S from 20pcs units, the result looks like that:

The right channel a bit worse:

 

[soundnerdie]

Very nice! Was it measured using the load board directly into ADCiso? Which input range did you use? 1.7Vrms? Thanks in advance.

 

[IVX]

load_board, of course, 9039S 0dbfs 3.5Vrms Cosmos APU->Cosmos ADC.

 

[capslock]

That is insanely good! What was the load on the outputs?

 

[IVX]

Only APU's 10k. Loaded at 0dbfs 32ohm all units are close to -124db of THD+N(380mW), -123db at 16ohm and -.5dbfs(680mW), and -121db at 12ohm -1.5dbfs(715mW).

 

[tombo56]

BTW, yesterday I selected one of the best THD 9039S from 20pcs units, the result looks like that:

OK, that's the best one. What is distortion for an average one?
If we want to use this as a low THD signal generator, would it be possible to order one with guaranteed lowest THD?

 

[IVX]

nearly all units have at least one channel with all harmonics <=-150db with MCLK 24MHz @96kHz of Fs.
For example, I keeping 2 more samples selected yesterday, they are worse than the best one for the THD but a bit better for THD+N, or one channel worse THD etc. Not too much different. Anyway, I recommend to recalibrate the DAC with APU on your side. THD+N is more stable, all 20 samples tested have it from -125.5-126.1db.

 

[capslock]

We've had the discussion about how an integrated DAC / ADC Cosmos combo with a common clock would be great for doing correlated analysis. How about a common supply voltage and above all, a common voltage reference? I haven't checked what kind of references ES9822 and ES9039S take, but if there was a way to use a common reference, this should greatly enhance LF S/N for analysis of analog circuits.

 

[IVX]

a common reference, this should greatly enhance LF S/N for analysis of analog circuits.

I don't think so.