Best spec ADC Chip currently.... ??

[DonH56]

My Earthworks M30 measurement mic is rated to 30 kHz, and there is a 50 kHz M50 version (I think @amirm has that one; he has all the best toys!) It is true many stage and recording mics are only specified to 20 kHz though many go well beyond that. Like speakers with tweeters that extend to 40 kHz or more, often the goal is to ensure that sounds up to or a little past 20 kHz are captured cleanly without distortion, so they have bandwidth that extends well beyond that even though the spec is only 20 kHz for most of them. I have used several mics like the Shure SM81 and some big AKG condensers (C414 family) up to about 25 kHz or so.

If this ADC is not suitable for your purposes without NIST or similar certification, then a search for a different ADC is in order.

 

[IVX]

but he isnt able to verivy to any international standard. He compares his measurement to his AP 27xx analyzer and nobody knw about he calibration of his unit

I built a lot of notch-filters and ultralow distortion DAC+LPF combo, and "Victor's oscillators" to confirm the lowest THD/THD+N levels, which are unavailable with any of APs at all. However, I did that not for Cosmos ADC with the best THD+N just -125db@1kHz and the best THD -140db that's quite trivial. I prepared that for the new coming Cosmos APU device with -160db 3rd harmonic at 1kHz, and at <-130db at 10kHz. Cosmos APU I have no idea how to calibrate indeed due to crazy low harmonics level, and THD+N <-131db.

 

[thulle]

G-K showed up on the E1DA discord and started ranting a few days back about anything and everything, from how the separate heating of the DAC-chip strains the PCB and solder connection to how people without his extensive education shouldn't be allowed to do measurements since they don't know what they're talking about.

I could see some points in what they were saying, it's basic stuff that can be found by any uneducated person trying to learn from all the awesome forum resources available. But the unprovoked ranting, and failure to explain what the intended goal and searched-for result of all the complaints are (should Ivan stop selling his ADC? Require a certificate of education with every order?), and instead just repeating the complaints a few times just made me relieved that they started getting banned from channels where they were just too off-topic and obnoxious.

Now he comes here to pick a fight with the moderator who banned him.

I'm one of the uneducated, unlicensed measurers that just want a nice way on top of listening to verify I don't have any major issues in my audio path. Even without G-K's "Mil spec" background I'm WELL AWARE that electronics have tradeoffs and that I'm not buying an AP measurement solution.

 

[IVX]

thulle, actually, I did not ban that guy. In my opinion, he can say what he thinks about calibration or whatever related to the "audio-theory" in our dedicated discord channel #audio-theory. But our moderator I believe better can see the situation and he decided to ban G-K.
PS: If someone still believes that AP offers some unbeatable distortions level measurement, take a look at that PDF: https://www.nanovolt.ch/resources/l.../pdf/low_distortion_oscillator_comparison.pdf
As you guys can see, AP doesn't guarantee even -140db(as I remember official specs <-130db) level of harmonics, and what APx555 owners "measure" at <-140db is pretty much the world of pink unicorns I don't saying if APx555b is junk, no, I would say it is 5-10x times overpriced.

 

[thulle]

@IVX I didn't say you did, G-X named the person who did a few pages back. I considered tagging G-X in the audio-theory channel to discuss the more interesting parts of the rants, but got too annoyed before reaching the end of it

 

[JimiPegorino]

Forgive me for asking but can the COSMOS ADC be used to record an effects/mastering chain? Looking for a cheap ADC for this purpose (I think just like the original OP). Thanks!

 

[thulle]

Forgive me for asking but can the COSMOS ADC be used to record an effects/mastering chain? Looking for a cheap ADC for this purpose (I think just like the original OP). Thanks!

It's one of the listed uses @ the Cosmos ADC homepage

 

[G-K]

Maybe l have a different understanding of calibration and there is a different meaning of the word calibration in German language. For me when a device is calibrated, the source signal must be calibrated to a tracaeble international standard and must be better than the device to be calibrated. The device to be calibrated will be checked if it fulfills the specification of the manufactured device. Of course two calibrated devices can still show different test resuluts. I measured COSMOS ADC at input range of 3.5V and l compared the test results with an AK5572 based ADC with almost the same input sensitivity (difference 0.5dB),both ADCs were connected in parallel to avoid differences in measuring results due to low input impedance of COSMOS ADC and Cosmos ADC shows higher THD than AK5572 ADC at input level of -10dBV. I gave such measuring results to the COSMOS discord channel and never got an answer what the real resaon for that is. From my side of view with low input level you get better THD measuring with AK5572 but with high levels you get better results with COSMOS ADC. I never said that COSMOS ADC is a bad ADC, but it depends of what you measure and you should know what you measure. A note to dbA measuring. When we use the A-weighting filter, we limit the frequency range and adjust to human hearing range and sensitivity. So why should we measure non hearable frequency above to 20kHz with sampling rates higher than 41kS/s? In short, all we want to measure with a high sampling rate ADC with high dynamic range and low noise, depends of what we need to measure and not for music only

 

[G-K]

Something out of topic here. I talked about test according MIL-STD. Some commercial companies claim to test their units, like laptops, according MIL-STD 810 and people think they geta quality unit, but the companies not even tell you according what Test procedure they do it and they dont tell you too that they dont fulfill all environmental conditions according MIL-STD 810, not to talk to all EMC tests required what are not related to MIL-STD 810. Testing to standards is time consument and make units much more expensive, same as calibration of test equipment. I know, hobbyists dont care about that

 

[G-K]

@IVX, The COSMOS ADC has CE lable. Can you tell to what CE standards COSMOS ADC is tested and can you provide the test results from an independent test facility?

 

[mansr]

Dude, what's your problem?

 

[Bamboszek]

@IVX, The COSMOS ADC has CE lable. Can you tell to what CE standards COSMOS ADC is tested and can you provide the test results from an independent test facility?

If you are so strict about certification, that you probably should know that CE is self-certified and doesn't require any independent lab (with some rare exception).

 

[IVX]

Bamboszek, I have that EMC report, it is a boring PDF, no idea what G-K gonna see there.

 

[G-K]

@IVX Thankyou for the document.

 

[G-K]

Here my measurement for the comparsion of COSMOS ADC vs AK5572 ADC. My problem is that l see higher distortion with COSMOS ADC with the identical input signal.

 

[freemansteve]

It's hard to get a DAC that isn't really good these days, as used as a component by any decent audio manufacturer...

You may well be able to hear differences, but it's not easy to tell which is best - in terms of overall audio experience in a system, or in terms of sensible value of money - just from reading specs or looking at plots. It's basically a religious thing, beyond a certain point...

 

[IVX]

Guys, as I see some of you still didn't read my short description for the Cosmos ADC https://e1dashz.wixsite.com/index/cosmos-adc

As I said, each ADC individually THD compensated at [email protected]. If you need to compensate distortions at another level, you have Cosmos_Tweak app for that. In my opinion, -.5dbfs is the best level to measure THD+N, because at -20dbfs THD+N anyway drops -100db or worse. ES9822 is pretty much the same level ADC as Ti or AKM has but 9822 has DSP able to correct typical distortions rise at 0..-10dbfs, and that simple fact lets us use Cosmos ADC as a THD/THD+N meter with down to THD+N -125db for some grade-0 units.

 

[G-K]

@IVX, I am sorry, but at levels <-10dBFs l can't improove the THD with the Tweak Cosmos application.

 

[IVX]

I can reduce 3rd(in that particular ADC H3 coefficient need to change from 12 to 1200) from -130db to -140db at -20dbfs input.

 

[KSTR]

Here my measurement for the comparsion of COSMOS ADC vs AK5572 ADC. My problem is that l see higher distortion with COSMOS ADC with the identical input signal.

I think what you are seeing here is that the ESS ADC, like most ADCs, is internally based on a DS-DAC core, and the DAC appears to be the same core as in the ESS DACs. And the ESS DACs have that unique spray of harmonics which is almost independent of level as long as the level is above ~30dBFS. While this spray of harmonics is usally just below the noise floor and thus does not impact THD+N, it still limits its usability a bit when one is to inspect the dirt buried in the noise floor -- which is best exposed by large FFTs at low sample rates, or alternatively by looking at the time-domain residual with large amounts of synchronous averaging (which sadly cannot be done with stock units of the Cosmos ADC as there is no means of syncing to a common clock).

I would agree that AK5572/74 is a stellar ADC when used at -20dB, basically flawless for THD-N. Close to 0dBFS the ESS is the better device for THD+N.