This is a 10s video of measurement THD+N E1DA Cosmos ADC residual THD+N -121/-123db 2ch/1ch mode at -.2dbfs level. - YouTube
syn08,
who knows, maybe you are a lucky mortal that did not see ESS legendary awful datasheets:
Best spec ADC Chip currently.... ?? | Page 2 | Audio Science Review (ASR) Forum
who knows, maybe you are a lucky mortal that did not see ESS legendary awful datasheets:
Best spec ADC Chip currently.... ?? | Page 2 | Audio Science Review (ASR) Forum
I tried several times to get some ESS chips from the NA representative and I was told my money is not good, since I am not going to eventually place a large order (whatever that means). Ismosys has no stock, and no idea when they will get them.
So apparently I will have to wait until Mouser will start distributing the chips, which could be tomorrow or two years from now, who knows?
About the ESS data sheets, I am aware of the mess, the DACs data sheets are nowhere better.
So apparently I will have to wait until Mouser will start distributing the chips, which could be tomorrow or two years from now, who knows?
About the ESS data sheets, I am aware of the mess, the DACs data sheets are nowhere better.
Ok, I tried again to place an order with Ismosys, this time it went through (no more pointing to the NA distributor pricks as before) and they are waiting for the payment to their bank account (including shipping), in GBP. Will go tomorrow to my bank to do the payment.
In the email, Ismosys says the order will be fulfilled by Astute Electronics Ltd?
I was not asked about signing any NDA, perhaps they will ask after I make the payment?
And how do I get the datasheet, is it already included?
In the email, Ismosys says the order will be fulfilled by Astute Electronics Ltd?
I was not asked about signing any NDA, perhaps they will ask after I make the payment?
And how do I get the datasheet, is it already included?
Here you go Cosmos ADC THD vs level sweep - YouTube
As you can see, ES9822 doesn't really need to keep input level <-6-10db because the ADC is quite linear up to 0dbfs(that's why you can measure THD+N as well!), so I recommend using it with that level, and I do a calibration(compensate 2nd, 3rd, and 5th harmonics) for the level -.5dbfs.
As you can see, ES9822 doesn't really need to keep input level <-6-10db because the ADC is quite linear up to 0dbfs(that's why you can measure THD+N as well!), so I recommend using it with that level, and I do a calibration(compensate 2nd, 3rd, and 5th harmonics) for the level -.5dbfs.
Last edited:
Ok, so where do I apply for the data sheet? I applied on the Ismosys web site, is that all I have to do? I was able to make the payment for the chips using my bank's online services, so now I'm waiting for the Ismosys confirmation.
Last edited:
Unbelievable, on-screen data update rate crazy fast.
Check this out:
MCP356X Linearity test. - YouTube
It looks so slow, 10 times or even more? How it could be? ADC sampling rate is actually higher, 62.5 kHz vs 48 kHz. FFT-32k takes 0.5 sec just to fill up an array. Certainly, I can use smaller size, 8.192k or 4.096k and accelerate "visual" responsiveness, but I 'd lost noise flour. My ADC shows about -132dB ground line, vs -160dB . To get -28dB down, FFT has to be enormously bigger, 20x times or so, up to 512k.
What I'm missing? Same sampling rate with PCM4202:
pcm4202 62 5kHZ - YouTube
Video with jumping magnitude by 1dB/step also shows quite fast settling time, but I doubt your ADC is running at 5 Msps?
Check this out:
MCP356X Linearity test. - YouTube
It looks so slow, 10 times or even more? How it could be? ADC sampling rate is actually higher, 62.5 kHz vs 48 kHz. FFT-32k takes 0.5 sec just to fill up an array. Certainly, I can use smaller size, 8.192k or 4.096k and accelerate "visual" responsiveness, but I 'd lost noise flour. My ADC shows about -132dB ground line, vs -160dB . To get -28dB down, FFT has to be enormously bigger, 20x times or so, up to 512k.
What I'm missing? Same sampling rate with PCM4202:
pcm4202 62 5kHZ - YouTube
Video with jumping magnitude by 1dB/step also shows quite fast settling time, but I doubt your ADC is running at 5 Msps?
MagicianT I'm not sure if you asking me but my setup was:
AP analyzer generated digital sine 24/48k, next SPDIF to my PC, and play that sine with Cosmos DAC(that device includes switchable LPF 1k and 10k to produce ultra-clean 1k/10k sine). 10Vrms from Cosmos DAC goes to Cosmos ADC XLRs inputs, 32/48k ADC result goes to the PC back to make there FFT 64k-bins and to resend by SPDIF 24/48k ADC result to the AP.
AP analyzer generated digital sine 24/48k, next SPDIF to my PC, and play that sine with Cosmos DAC(that device includes switchable LPF 1k and 10k to produce ultra-clean 1k/10k sine). 10Vrms from Cosmos DAC goes to Cosmos ADC XLRs inputs, 32/48k ADC result goes to the PC back to make there FFT 64k-bins and to resend by SPDIF 24/48k ADC result to the AP.
I just did add the I2S connector footprint that was asked by PM 5 min ago 
PS: frontend resistors are Susumu RR1220PD and CSR02027, opamps OPA1602, power Monolithic Power Systems and Ti, USB bridge CT7601CR as a slave, ES9822 as master. The USB power has a soft-start, hence, ADC can run even with a smartphone if needed.
PS: frontend resistors are Susumu RR1220PD and CSR02027, opamps OPA1602, power Monolithic Power Systems and Ti, USB bridge CT7601CR as a slave, ES9822 as master. The USB power has a soft-start, hence, ADC can run even with a smartphone if needed.
Attachments
Last edited:
O'k, 64k FFT than.
1. Sampling with 48 kHz it takes 64 /48= 1.33 seconds to fill a data array. Using 50% overlap 0.67 seconds. All digital data shown on a video updated 10x faster. Is video played at 10x speed?
2. With 64k FFT noway to get -160dB noise flour. About -140 at the best.
So, the real world is fake anytime when we do not expect if something does possible?
Attachments
I did take a look at the ADS127L0, this is quite noisy ADC SNR 115db, if you'll check other ADCs(over 120db range) like PCM4222 or AK5578 you'll find much less than -140db noise-floor without Boltzmann constant trimming or elementary charge boosting in the flux-capacitor
