- 积分
- 244
- 在线时间
- 929 小时
- 最后登录
- 2023-7-5
- 阅读权限
- 50
- 精华
- 0
- UID
- 834496
- 帖子
- 234
- 精华
- 0
- 经验
- 244 点
- 金钱
- 218 ￥
- 注册时间
- 2017-6-28
|
楼主 |
发表于 2021-10-17 23:12
|
显示全部楼层
This text is the copy from the product page https://e1dashz.wixsite.com/index/cosmos-adc
Let me say you something about it.
I have to apologize for my non-native English, and non-literary style. I just want shortly and not too boring to describe what I know about that ADC.
A) We offered our 9038S/D DACs users to control the distortions performance of those DACs with the TWEAK9038 Android app.
B) Who says A must say B..we prepared an instrument to test that performance accurately.
You can calibrate your 9038S/D DAC for the minimum distortions at the particular headphones the DAC loaded, or measure non-E1DA DACs THD/THD+N to see how far was a “golden sample”, measured by some blogger, from your particular DAC ;) Oh, a lot of fun around the high-performance audio-analyzer on your palm! Especially if you are a perfectionist as I am. I remember I bought a Swiss spectrophotometer for my DSLR, both were the same price. So, there is the same situation regarding audio performance, I offer you to test and calibrate distortions of your E1DA DACs by yourself.
Of course, Cosmos ADC can be used as the finest vinyl ripping ADC or as a studio recording interface but I think pretty much any ADC may rip vinyl(just 70db of the 125db dynamic range needed) or record a mic's signal, where 16bits resolution, I guess is enough. In other words, Cosmos ADC is for sure breakthrough in audio measurement which becomes so attractively inexpensive and top-precise at the same time but regarding the vinyl ripping/studio recording, I don't think if you may notice the difference between an ADC with 10x times worse distortions VS Cosmos ADC.
I don't insist I'm right, fortunately, 10x times fewer distortions Cosmos ADC cost >10x times less than "top brand" ADCs, and you can buy it, try and report back in our discord channel about how I was right/wrong.
Cosmos ADC is a plain USB-C powered Analog to Digital Converter to PCM 32bits/44.1-384kHz with extremely low distortions and noise, which the first time in history makes the ADC suitable for the THD+N measurement at levels lower than audio analyzers $20000 price range. I agree, it sounds suspicious but what can we do with the tech progress? Cosmos ADC uses the ES9822Pro chip released by ESS Technology. Inc. less than 1 year ago. The chip is positioned as "The Sabre ES9822 PRO is the world’s highest performance 32-bit analog-to-digital (A/D) converter targeted for professional audio applications such as recording systems, mixer consoles and digital audio workstations (DAW), test equipment, instruments". Sounds too bright? Actually, it's a true statement, at least with formal specs like THD+N ESS chip overtakes competitors for 7-10db or so, and at the same time outperforms my old analyzer AP SYS2*** for 2db(typical for Cosmos ADC grade-B in mono mode).
Regarding the grading, each Cosmos ADC unit is individually calibrated for the minimal distortions at level -0.5dbfs 1kHz with the typical THD <0.00003% or to say shorter .3-.2PPM. The typical level of 2nd, 3rd, and other harmonics between -135.. -145db for all grades from A to C. Grades are only about the noise, and each unit has a grade mark on the left side of the front panel. In most cases, SNR/Dynamic_Range in the MONO mode is within 127+/-1db(A) it is grade-B Cosmos ADC. Less than 126db(A) means grade-C, from 128db(A) and better grade-A, 129db(A) and better grade-0(very rare grade, and probably redundant, just in case). The MONO mode improves the SNR/DR for 3db, so a grade-A unit typically has SNR 125db(A) in STEREO mode. MONO/STEREO modes are switchable by the UAC2 volume commands, for example, if your Windows volume slider position is less than 50%(0 to 49%) the device works as 2 channel ADC, if more than 50% it is a MONO ADC with summed Left+Right inputs mapped on the Left channel(Right input is always mapped to the Right channel). In the MONO mode, both channels need to be tied together L+ to R+, L- to L-.
The free FFT software list we recommend use is quite short:
1) Arta https://www.artalabs.hr/download.htm should be used with 20-20kHz BW limit as a fake mic-response file https://drive.google.com/file/d/ ... U/view?usp=sharing. It is a free app but to save files need to pay $100, I don't remember if I had ever needed to save something there. Arta correctly calculates SNR, THD, THD+N, but as I noticed incorrect IMD.
2) REW https://www.roomeqwizard.com/ free for a single-channel version, that's enough for most cases. Not all versions correctly work with THD+N but RC9 at least is ok. The known issue is ASIO only supported, ASIO4ALL works ok but up to 192K. Sometimes ASIO4ALL driver can occupy Windows 10 volume slider after use, which makes it impossible to change the MONO/STEREO mode of ADC. Another nice option for REW is FlexASIO with 32/384 https://github.com/dechamps/FlexASIO/releases
Most FFT software may correctly show the levels of the harmonics but incorrectly calculate THD+N etc be aware. Please watch the tutorial about Arta using with Cosmos and 9038S/D DACs https://youtu.be/eguKMNZ-XJw
Also, you have a free E1DA Cosmos Tweak Windows app. This app lets you control the distortions harmonics level, interpolation filters, and ADC's integrator settings. The harmonics control is similar to our TWEAK9038 Android app for DACs but without a volume dependency. All ADC units are calibrated before selling, so it is a good idea do not change Cosmos Tweak Windows app's options without copying calibrated values before. I.e. after you plug the ADC, run TweakCosmosQT.exe, and click on the [Connect&Read] button to see the calibrated values of THD compensation(THD ADC1/3: 2nd, 3rd as the Left ch and THD ADC2/4: 2nd, 3rd as the Right ch), and ADC_INT_SEL for ADC1-4(usually, all parameters there = 11). Please write down all those parameters or screen it shot to recall if needed. To make sure if calibrated values are correct, click the [Connect&Read] button again to see that the app is connected to the hardware, do not lose the calibration values! Typical 2nd compensation values are 0+/-5, 3rd from -7 to -15. Why you may need that app? In case, if you have a precise sine-wave source with surely ultra-low distortions(AP analyzer, custom -150db clean 1kHz oscillator, or Cosmos DAC), you can re-calibrate Cosmos ADC distortions. For example, AP analyzers require regular annual re-calibration, hence, such a precise ADC as Cosmos needs to be re-calibrated for THD as well. ADC_INT_SEL parameters may improve the SNR of ADC. The slider 2ch MODE/1ch MODE does switch STEREO/MONO mode of ADC, the same way as a volume slider. You can set any of the 8 interpolation filters you wish, the default one is Linear phase Apodizing. If you care about latency, your choice is Min phase filters. All parameters changing are memorized automatically.
Cosmos ADC has a hidden on its bottom side the DIP-switch attenuator for 8 ranges of sensitivity 1.7-10Vrms. Each range has a different input impedance due to the ADC has only bare FDA inputs no high-impedance buffers to keep distortions and noise specs as high as possible. That fact needs to keep in mind if you gonna measure some old vintage stuff with relatively high output impedance. Modern DACs and DACs+HPA as E1DA 9038S/D for sure no problem. Cosmos ADC has 2 FDA XLR inputs 1.7-10Vrms, and a tiny 2.5mm 43Vrms AUX input as a gift to DIYers who like to play with power amps(do not plug/unplug that jack when the amp is On!). All inputs are AC coupled(-2V to +12v DC allowed) Fully Differential aka balanced and could be used with balanced/unbalanced outputs in any combinations i.e. doesn't matter +IN to GND or -IN to GND or IN+ to the OUT+ or to OUT-, the same for IN- etc.
The rear panel contains a USB-C socket and the orange status LED to indicate the sample rate by LED's flashing frequency(48k 1 blink per 2S, 96k 1 blink per 1S, 192k 2 blinks per second etc). I saw on reddit.com complained about LED flashing. Anyone has "am I boomer?" psychological complex, who feels the LED blinking is too much 1960th style, please request in our discord channel for the Boomer_free Firmware with no LED flashing + how2reflash instructions - you are welcome )
LEDs on the front panel work as a "Signal" indication - green if the STEREO mode and the input > -30dbfs, or blue in the MONO mode. A "Clipping" red-LEDs indicate that the input level reached 0dbfs. i.e. 80s style as I believe ;)
Accessories:
1) Cosmos Load gold-plated board, is a dummy load to test headphones-outputs, balanced(DIP-switch 11&12 Off) and unbalanced(DIP-switch 11&12 On). Each of the two channels could be independently configured by the DIP-switch for 32 variants of the resistance value 10-300ohm + no load. The board has 2 plugs 3.5mm(if DUT has unbalanced headphones output) and 2.5mm(if balanced), you need to install the plug into a Device Under Test(DUT). Also, the board has 3.5mm and 2.5mm jacks, you can use these for real headphones as a load. In that case, the resistance DIP-switch needs to be set as no_load to measure the distortions on the particular cans. The balanced side has 4 outputs L+, L-, R+, R-, and unbalanced 4 outputs L+, L-, R+, R-, please use hook-clamps(red-POS and white-NEG Inputs) of the XLR probe cable to tie the Load board to the ADC. The GND crocodile sometimes makes sense to tie to the USB-C shield of a DUT.
2) to connect Cosmos ADC to the Load board we prepared the XLR probe cable. It is a 1/2m short a decent quality shielded cable(a lot of copper!) with two hook-clamps(red-POS and white-NEG Inputs) and one crocodile-clamp(GND) on the DUT end.
So far I have mentioned pretty much Pros only, let me talk about Cons too.
1) USB-C is a tiny modern socket with < 0.3mm width golden pins, UAC2 utilized in that ADC uses a bulk transfer protocol without bad-block repeat requests, unlike the data-storage transfer. Hence, if you gonna plug/unplug the USB-C cable everyday, you'll start to complain about an unreliable USB-C connection next 3 months. A good idea to use USBA2C cable and plug/unplug its A-end on the PC side.
2) Each ADC unit is individually calibrated at a particular level -.5dbfs, and 1kHz frequency. For the best precision of measurements, you also need to keep the input level close to that level. THD across the 0-30dbfs range will be within -120db=1PPM it is not bad but the best is -.5db and most of the samples have there .2PPM. The same regarding the frequency, Cosmos ADC calibrated at 1kHz because it is a standard THD+N test in the audio industry. If you need to measure 10kHz THD better to re-calibrate the ADC with a reference 10kHz sine source first(see above E1DA Cosmos Tweak app) or enjoy 1-1.5PPM of THD.
3) The calibration is performed at the particular ambient temperature 27C, if your environment is seriously colder/hotter the THD performance is not guaranteed. Every time the unit needs to be warmed up during 15-20min before use.
4) The front end of Cosmos ADC has a very wide frequency response up to 185-190kHz and using that in 44/48k Sampling Rate you may notice some aliasing artifacts due to DAC's HF noise not filtered out well. Try to switch SR to 96-384k to make sure if some strange FFT peak you found is real.
5) to preserve the original ESS ADC chip performance as high as possible and keeping the cost under tight control, the ADC utilized a direct FDA front end with different input impedance across the gain ranges. It makes voltage drops at the DUT outputs if one has a relatively high output impedance and minimal ADC input range selected(1.7V, see the specs), I did mention that above and have just repeat it again.
|
|
发表于 2021-10-14 15:47
