• 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!

Best spec ADC Chip currently.... ??

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Grooved, to be honest, I don't care about the studio-production market, there is brand-name needed but not tech specs. If you are The Name, you can sell the same model for 15 years for $2000-4000/pcs with THD+N -115db, and Crosstalk -115db as a TOP of technology. I have sold about 2000pcs DACs with THD calibration ability, Cosmos ADC for that. I think if someone in a modern studio aren't stupid, he'll buy my Cosmos for 10% of The Name price, maybe 10-50pcs of ADCs would be sold there, why I should care?
 

Vuki

Senior Member
Joined
Apr 8, 2018
Messages
341
Likes
392
Location
Zagreb, Croatia
@IVX, would you include RIAA filter in Cosmos? Few years ago I needed ADC to be able to connect turntable and fm tuner to my preampless system and made small ADC based on ES9102+OPA1632+arduino control. It works OK. But me being a dentist and you being EE, I guess your Cosmos would be in a different league...
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
Grooved, to be honest, I don't care about the studio-production market, there is brand-name needed but not tech specs. If you are The Name, you can sell the same model for 15 years for $2000-4000/pcs with THD+N -115db, and Crosstalk -115db as a TOP of technology. I have sold about 2000pcs DACs with THD calibration ability, Cosmos ADC for that. I think if someone in a modern studio aren't stupid, he'll buy my Cosmos for 10% of The Name price, maybe 10-50pcs of ADCs would be sold there, why I should care?
Sure if it's for 10-50, you should not care. It's just that I imagined it would be a far higher number than 10-50 if it's really a great device, even without the big Name, especially for an ADC (you more often needs more than two channels for outputs but it's not always the same case for ADC)
 
Last edited:

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Grooved I did post all preliminary specs already, what do you think isn't great over there? Dynamic range 127/124db(A), THD+N -121/-118db mono/stereo? Or maybe crosstalk@1kHz -135-140db is not enough? What I should improve?
Vuki I have no such plans yet. You can apply RIAA in any biquad later or even offline to the audio file.
 

Vuki

Senior Member
Joined
Apr 8, 2018
Messages
341
Likes
392
Location
Zagreb, Croatia
Grooved I did post all preliminary specs already, what do you think isn't great over there? Dynamic range 127/124db(A), THD+N -121/-118db mono/stereo? Or maybe crosstalk@1kHz -135-140db is not enough? What I should improve?
Vuki I have no such plans yet. You can apply RIAA in any biquad later or even offline to the audio file.
Well, I'm using it with simple low noise linear preamp like a phono preamp to listen to phono in real time.
 

Vuki

Senior Member
Joined
Apr 8, 2018
Messages
341
Likes
392
Location
Zagreb, Croatia
Vuki to be honest I didn't get the application info from ESS about how to set up ES9822 internal DSP. They gave me some information only for their Windows control app, I did ask about the same but more practical one from the MCU programming point of view - no reply.
I got the es9102 with a datasheet as a gift when it was already obsolete so no contact or support from ess. Fortunately it turned out alright. Got a button to select riaa (works only @48khz), 48, 96 or 192khz. :)
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
Grooved I did post all preliminary specs already, what do you think isn't great over there? Dynamic range 127/124db(A), THD+N -121/-118db mono/stereo? Or maybe crosstalk@1kHz -135-140db is not enough? What I should improve?
I never said something like that, your numbers are great ;-)
I only thought a lot of people should be interested, but that it could interest even more people with a feature like clock plug in case of a need to use several devices in sync, or an SPDIF or AES output,... and wanted to know if it's something you may add to the Pro version, or on this one
(because I understood that you may release two versions, but maybe I misread it)
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,013
Location
Berlin, Germany
@KSTR
An interesting approach!
I tested the AK4493 in my own design some time ago and got a THD of -125dB at -6dBFS. If that can be pushed down to around -140dB it would be fantastic. And by using 2 devices per channel the dynamic range should still be around 123dB.
Looking forward to learn more about the details of your design!
Short update.
Once we go into the -120dB range funny things come to light, in this case again the omnipresent cancelling. For example, in the AK4490-equipped DAC, it turned out at high levels H2 is way different depending on where I tap of the signal. Right at the chip it is unexpectedly way higher (10...20dB!) than after RME's lowpass filter and that is somewhat consistent across the four channels I have for testing. I think that can be only the result of cancelling in that OpAmp stage, I have no other explaination, except the filter caps maybe (I have yet to swap filter stage inputs pins to see if the assumption holds but soldering around in the RME is no fun and I don't want to overdo it, risking damage). Test is also due for the AK4493.

It's a real pity that AKM is shut down and I have little hope they will recover quickly, with chips unavailable for many months to come. Making a test board for 4493 makes little sense atm.

What I can say now, for both chips, is that one can push or pull the DC voltage at the Vcm pins a little and that acts as a distortion trim (looks reasonable from the SC-DAC core section shown for AK439x), odd orders can be greatly reduced at levels only few dB below full-scale. In interplay with digital DC-bias a sweet spot might be found with low odd orders and the remaining dominant H2 and H4 having close to identical levels between channels but opposite phase so cancelling would be quite effective.
... but it didn't work out fully because of the above. I also tried unloading the chip outputs with a positive feedback "bootstrap" circuit (the balanced output from the filter stage allows for a very simple circuit) which should also help reduce distortion at high levels a bit but again I think the not so very promising results were skewed as I had tried all this inspecting the signal after the filter stage.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Tks Sure, preparing for the production, so far Light version with DIP-switch attenuator 8 gain steps, just to keep the price and performance as good as possible. $130/pcs or less(depends on a $ilicon parts prices), Jul-August release expected. The most disputable point is quite low input impedance due to no input buffers(680ohm @ 1.7Vrms, there is only USB bridge, ADC, and inverted opamp stage for balanced inputs, pretty much the same as ESS demoboard). For my 9038S/D DACs outputs, such an input impedance isn't an issue at all, maybe some DACs outputs will not really like 1300ohm at 4Vrms or 2600ohm at 8Vrms However, I don't know any other DAC with the ability of a calibrating of 2nd/3rd harmonics so why I should care? I positioning that ADC as an audio measurement instrument for my DACs, if anyone wants to use that for a studio recording(I see requests in my discord about it), please but RME or Hilo or whatever, because a production studio doesn't need a cutting edge performance at all, rather need the brand which I have no.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,494
That's good to hear. Yeah a bit weird people want some pro interface when that requires proper support channels and all that. How feasible would a general purpose interface be btw? Or is that also not possible without massive scale? I'm wondering something super basic like a Schiit Hel sort of thing?

Btw, if you get a chance, show us some images :-] would be cool to see what it would generally look like.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Tks, not sure what do you mean regarding Schiit Hel. I did take a look, this is just a DAC. Cosmos ADC is the simple USB(single Type-C, no extra power required) analog2digital convertor with 2 XLRs(1.7Vrms..10Vrms 8 steps DIP-switch) and one 2.5mm jack inputs(AUX 20..30Vrms to measure power amps). Mono/Stereo mode is switchable by UAC2 Volume(>50% of Windows volume slider means MONO, <50% STEREO). Mono mode gives 3db advantages for SNR, Dynamic Range(DR), and THD+N(aka sinad), for instance, DR in 2ch mode 124db(A), 1ch 127db(A), residual THD+N at 2ch mode -118-119db, 1ch <-121db etc. Each channel has Signal LED(Green for stereo, Blue for mono mode), and Clipping LED(Red). Max Fs 384kHz(frequency response +0.1/-3db 5Hz-190kHz), max bit depth 32bits PCM. Crosstalk -140db@1kHz.
PS: also attached Cosmos Analog Processing Unit, which contains Low Noise Amp(LNA balanced input with CMRR -120db, 140nVrms(A) 40/60db to measure DR/SNR of any imaginable monsters DACs), and notch filter 1/10kHz with residual THD+N -128db@1kHz, harmonics level -145db at 6Vrms input(-135db at 10Vrms, -155db or so at 1-2Vrms). Cosmos APU will be $75. In fact, calibrated Cosmos APU could be used with any laptop's audio inputs to measure so low levels of distortions. Super-duper ADC isn't required i.e. you can not measure the noise of Meizu HiFi DAC(~550nV(A)) with APx555b(~700nV(A)) because of the huge residual noise of AP analyzer but you can buy Cosmos APU for $75 and measure 550nV no problem ;)
Cosmos_ADC.jpg
Cosmos_ADC_trans.jpg
Cosmos_APU.jpg
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,494
The Hel is also an ADC (basic 3.5mm mic input). But wow, that looks great dude.

It would be funny seeing the AP measuring something, that can measure better than the AP :p

Can't wait to get one of these though in all honesty.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
The 3rd device in the Cosmos line is Cosmos DAC, dual ES9038Q2M based, in the same case as ADC and APU. DR 130db(A) THD+N@1kHz <-124db or .6PPM 0.00006%. 2 XLRs and 3.5mm jack(HPA) outputs. Each XLR could be switched to be normal DAC balanced output or DAC+LPF 1/10kHz balanced output 10Vrms to use such harmonics-filtered output as an ultra-pure generator for power amps measurements. DAC+LPF output to Cosmos APU next to Cosmos ADC shows -128db THD+N@1kHz. Price is truly TBD but within $150-170 limits in my mind.
PS: the notch had -20db reduction ratio, so the reading -104db needs to be corrected to -20db, hence, THD+N <-124db, the same about 0.0006% reading needs to be divided by factor 10.
 
Last edited:

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
I forgot to mention that Cosmos APU has switchable passive LRC 4ord LPF 22k before LNA in case if needs to cut nonaudio noise-content i.e. BW limited at the input and such a problem that typical for AP analyzers(the input BW isn't limited, if you'll measure something contains a tiny amount of HF disturbances - PC/classD/SMPS noise, AP may go mad and show unrealistic results) near to fully eliminated.
Cosmos ADC seems quite ok for the "jitter 12kHz test", please see the attached FFT 1M 15x AVG of 9038SG3(S3000), the only single sideband over -160db is noticeable.
2021-07-08_21-28-48.jpg
 
Last edited:

MC_RME

Addicted to Fun and Learning
Technical Expert
Audio Company
Joined
May 15, 2019
Messages
854
Likes
3,564
Cosmos ADC and APU look like cost effective solutions for mobile measurement (in my case), and a useful addition even for owners of the AP. So where can I order these? You maintain a pre-order or waiting list?
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
MC_RME, no pre-order, just gonna produce 1KU of each.

The initial Cosmos ADC(I gonna release that later as Cosmos ADC Pro) with 8 steps MOS-attenuator is also ready and very slightly worse vs lite one for DR/SNR(about .3db), the THD compensation in fact the same, hence the MOS-attenuator is true WOW and has some performance room for even better ADC than ES9822. Why did I decide to go first with the simplified Cosmos ADC? A few reasons are there:

a) Input buffers, differential stages are needed to be powered with +/-16Vdc to be able to work with up to 10Vrms input. OPA1612 has about 7.3mA of consumption idle current, 7 OPA1612 vs 3 in the lite version. All stages are 600ohm or less loaded for the lowest noise performance. 10Vrms at 600ohm or less in 4 nodes need about 700mW of power! The total number of SMPS regulated power rails reaches 4, lite one only 2. Finally Pro draws 1.5A at 384k mode 10Vrms input when the lite one just about 200mA. For USB3 specs 1.5A is still within the limit but many USB2 will be overloaded by Cosmos ADC Pro.
b) The B.O.M. cost($ilicon parts bloody expensive) and calibration time. The lite one needs only one THD compensation across all gain ranges because only different MELF0207 resistors(a huge res case with virtually zero VCR) will be switched by the DIP-switch. Pro will need 8 slightly different THD compensation coefficients for each range.
c) Cosmos ADC + Cosmos APU with a comparable cost will provide a lot more than just Cosmos ADC Pro with nice UAC2 gain control + 20Kohm input impedance.
 
Top Bottom