• 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!

E1DA Cosmos APU - Teardown and personal thoughts and measurements

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
Hello all,

I recently received a Cosmos APU from Ivan (@IVX) to have it teared down and, why not, to have its guts tested as much as possible. Main product page is https://e1dashz.wixsite.com/index/cosmos-apu from where we can get lots of technical details about this interesting audio electronics tool, while https://pt.aliexpress.com/item/1005004336488518.html has about the same content plus a great feature called “Buy Now”. :)

The smell of wood when I open up the case is very pleasant and it makes me take a deep breath before taking out the small “beast” from the case.
IMG_0755.jpg


IMG_0756.jpg

E1DA Cosmos APU - luxury wood packaging, like a small bottle of old whiskey

The idea of creating such an APU is, most likely, related to finding a cheap and viable way to upgrade any ADC, even your computer’s sound-card, and transforming it into a “home audio measuring device” by extending its resolution and dynamic and expanding its audio measurements capabilities with 30 dB for SINAD and with a selectable 34 dB / 60 dB for dynamic.

On top of that, the addition of +48 V phantom power will make the Cosmos APU, chained with any ADC, in my case it’ll be Cosmos ADC, a top notch combo for any compatible condenser or dynamic microphone out there. Basically, there haven’t been manufactured yet an audio interface having a better performance than the above combo, and I’m referring now strictly to THD+N and dynamics. In terms of gain, the +60 dB pre-amplification will match the max. gain of my Motu M4 and will even outperform Focusrite Clarret’s +57 dB of gain, meaning that Cosmos APU + ADC combo can be successfully used with low-sensitivity mics like AKG D5 and Sure SM7B.

However, worth mentioning that Cosmos APU has no adjustable gain, just the two fixed gains selectable from the +34 dB/+60 dB switch, so perfectly matching the proper gain value might be challenging. In my case I was using +60 dB for the AKG D5 S mic and +34 dB for the Superlux ECM999 and SE Electronics X1S mics, for the latest two needing phantom power activated as well. It wasn’t a perfectly matched gain for the last two mics (it would’ve been more helpful few more dBs, but given that the audio recordings were not having audible noise induced by the mics themselves, I am able to post-process the audio tracks in no time by simply increasing the audio levels.

APU’s basic schematic would be the one from below:
Cosmos_APU_Notch_simplified_diagram.png

The first stage (from right) contains a +20 dB gain stage, followed by the adjustable notch filters (@ 1KHz or @10 KHz), then the output stage with a selectable 0 dB/+6 dB gain. There’s also a +34 dB/+60 dB pre-amplification stage that can be used for measuring the background noise or DUT’s dynamic with any ADC you have around.

The COSMOS APU is using the "newly arrived" op-amp RT6863S, possible related to https://www.scribd.com/document/494918453/RT6863-DS-Rev-E or to https://www.semiteh.com/admin/upload/semiteh1-123171535.pdf, but also the well-known very low-noise OPA1612 op-amps. If the first op-amp is used in the output stage, near the the 0/+6 dB switch, the two OPA1612 op-amps are used for the Notch filters and for the pre-amplification stage (input buffer and composite differential amplifier).

The internal preamplifier for the +34/+60 dB input stage is built around the low-noise & high performance integrated circuit THAT1510. This input stage is the one that can also be used as microphone inputs too, not just for audio measurements and THAT integrated circuits are also used by many manufacturers of audio interfaces, with very good results, MOTU and Focusrite being couple of them.
IMG_0758.jpg

Preamp and Notch inputs
(notice the red rubber used as protection for the phantom power, to avoid accidental switching)


IMG_0759.jpg

USB-C connector, so plenty of cables can be found on the market to power the APU

As you can see, the external design is pretty similar with the one used with Cosmos ADC, having an identical case, just the side plates are different due to the inputs and outputs plugs.

Internally, we can see components of a very high quality, perfectly soldered, with a silkscreen that explains every component and switch, even the power rails too.​

IMG_0764.jpg

Top of the PCB
(notice the ground plane, the nice layout, silkscreen and the big caps that lowers the ripple)


IMG_0763_.jpg

Bottom of the PCB
(notice the properly soldered screws to the GND)​


It came to me uncalibrated, meaning that the notch filters have not been adjusted at all, nor the frequency, nor the amplitude (gain). I had it calibrated easily, although it took almost an hour to realise what kind of sweep file I need to play from REW (see couple of paragraphs below).

Calibration schematic is presented below, so all you need is an isolated screwdriver, so your internal resistance to the ground to not affect too much the values of the resistors you’re going to adjust:
Calibration_schematic.png


I had initially adjusted F1 and G1, corresponding to the frequency and the for the 1 KHz notch frequency, then I adjusted the F10 and G10 adjustable resistors to match the -30 dB notch filter reduction for the fundamental frequency of 10 KHz.
IMG_0765.jpg

We can easily notice the G1, F1 and G10, F10 (bottom of image, a bit out of focus due to DoF) variable resistors​


I’ve used REW as signal generator and I've chosen the Multitone generator with a 1 Hz step, like in the below screenshot. This way was easier to me to perform the calibration of the two notch frequencies properly. I also zoomed in to double check if the main frequency was exactly 1 KHz or 10 KHz.
REW_Multitone_1Hz.png

REW - Multitone generator (1 Hz step)


E1DA_Cosmos_APU-Calibration-1KHz_.jpg

Top of the graph hits the -39 dB while bottom @1 KHz hits the -69 dB, so a difference of 30 dB after a proper calibration of the notch filter


E1DA_Cosmos_APU-Calibration-10KHz_.jpg

30 dB notch filter calibrated @10 KHz

To check myself if I did a good job with the calibration, I did some frequency response tests, this time by using ARTA. See the two below graphs:​
E1DA_Cosmos-Notch_1KHz_resized.png

@1 KHz notch filter

E1DA_Cosmos-Notch_10KHz_resized.png

@10 KHz notch filter

Ivan has a calibration file that can be used when performing measurements (do not use this file when calibrating the APU!). It can be loaded in REW or ARTA to perfectly match APU’s internal response. More details about this calibration file and how to create such a file by yourself can be found in @Archimago article from here: http://archimago.blogspot.com/2022/05/early-look-cosmos-apu-high-performance.html/.

To perform the notch calibration, I powered the APU from the Schiit Wyrd low-noise USB power supply. After the calibration was done, I measured the two voltage rails too, just to make sure the APU has really low-noise power rails inside. The AC ripple & noise is almost existent, even when I was powering the APU from a rather cheap XTAR PB2S power bank that most likely has a cheap and noisy buck converter inside. Same flat / no ripple can be seen on the +48 V phantom power too.
E1DA_APU-48VDC.jpgScreen Shot 2022-09-19 at 11.00.20 PM.pngScreen Shot 2022-09-19 at 10.55.16 PM.png


“Responsible” by these low-noise rails is the TPS65130 - Dual Positive and Negative Output DC-DC Converter made by Texas Instrument. So basically, it accepts +5 V on the input and makes +/- 15.6V to power up the components from inside this APU. Of course, the nearby capacitors are definitely helping out in lowering the AC ripple & noise too.

What can Cosmos APU measure and how? Well, given the +34 dB/+60 dB internal pre-amplification, Cosmos APU can increase an external ADC measurement capabilities with +34 / 60 dB when measuring the dynamic of an audio device. Same applies when measuring the THD+N or the SINAD, by adding to the final measurements +30 dB due to the notch filter built-in and properly adjusted by the customer or by Ivan, depending on the APU bundle you’ve purchased.
ADC_vs_APU+ADC.png

Cosmos ADC vs. Cosmos APU + Cosmos ADC

Given the existing measurements performed already by @Archimago (http://archimago.blogspot.com/2022/05/early-look-cosmos-apu-high-performance.html) and @WolfX-700 (https://www-l7audiolab-com.translat...l=auto&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=wapp) it makes no sense for me now to upload my own measurements here, but I got similar figures as they did, so this APU really shines and makes a great addition to anyone needing a cheap device to "upgrade" an existing ADC.

I would like to thank again to Ivan for sending me this impressive piece of audio measurement gear that turns any ADC in a real measurement device. I also much appreciate the communication with Ivan, although I am aware that he prefers Discord channels instead of messaging of other kind. :)

I strongly recommend the E1DA Cosmos APU to anyone that wants to raise the bar in audio measurements, to those needing a cheap way to measure audio stuff with an existing ADC, but also to those needing a very low-noise preamp for their microphones.

Oh, in case you haven’t noticed the Coat of arms written on the back of the PCB and visible on the back of the case too, I’m telling you that it’s there. I wish you all peace and God help us to stop the war!
IMG_0760.jpg

P.S.: If you decide to purchase the Cosmos APU, then don’t forget to add to cart the dedicated audio cables like https://www.aliexpress.com/item/1005004389528210.html, otherwise you’ll need to DIY or procure jack-stereo to TRS plugs, case when you’ll need to use a multimeter to check if these cables or plugs/adapters are properly done (GND should not short-circuit with any of the other two hot wires!).​
 

Attachments

  • IMG_0769.jpg
    IMG_0769.jpg
    321.2 KB · Views: 313
  • IMG_0768_.jpg
    IMG_0768_.jpg
    316.1 KB · Views: 270
  • IMG_0767_.jpg
    IMG_0767_.jpg
    226.4 KB · Views: 321
Last edited:

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
I have been using a standard 2T notch circuit for 8 years. Is it the same thing, or am I missing something?

1663964499094.png
 

JW001

Member
Joined
Mar 22, 2019
Messages
84
Likes
95
Location
East Coast
https://e1dashz.wixsite.com/index/cosmos-apu :

"Please, before you decide to buy Cosmos APU, make sure that you completely understand what is it. Analog Processing Unit is the first in its kind device on the market, hence, you need to understand do you really want:
a) to do the measurement of your DAC/AMP distortions@1kHz/10kHz at <-150/-130db harmonics level, or THD+N@1kHz down to -132db. In that case, you need the Notch part of Cosmos APU.
b) measuring your DAC/AMP DR/SNR with the residual APU's noise 130nVrms(A) with the Preamp of Cosmos APU.
c) to rip vinyl records with the Preamp of Cosmos APU + Cosmos ADC with built-in DSP RIAA EQing. Please, check Cosmos ADC page to find the CosmosAdc_v13_RIAA.hex.
d) to use a Pro condenser mic with 48V phantom power with the Preamp of Cosmos APU."
 

GXAlan

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,868
Likes
5,954
Thanks for this review. For me it is just one of those days I simply do not understand one thing about what this thing, (APU ?) is is and what it can do…

This is just for testing gear at 1kHz and 10 kHz test tones. It improves the ability for the E1DA Cosmos which is a game changer for hobbyists to be able to test gear at a level competitive to the APx555
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,054
So way back in the old days, you would shoot 1 khz thru your gear, and have a unit to notch out the 1 khz, and you could see what was left.

This is the modern version of that. Why is it better than notching it out digitally? Because you notch it out in this box, and then feed it to a measuring ADC. Since most of the fundamental is gone, you can improve how good your measurements are of what is left by in this case 30 db or so vs using DSP only. Some DACs are now so close to the limits of good ADCs you are never sure how much is the ADC and how much is the DAC. This lets you see lower effective limits of the ADC to see how much is the DAC by pre-filtering the signal seen by the ADC.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
However, worth mentioning that Cosmos APU has no adjustable gain, just the two fixed gains selectable from the +34 dB/+60 dB switch, so perfectly matching the proper gain value might be challenging.
maybe, however, Cosmos ADC has a gain range 1.7-10Vrms so you have an additional 15db to adjust it ;) 34db and 60db is compromising between 3 different tasks:
a) DACs, amps, LDOs noise measurement.
Example:
AES17 Dynamic Range test performs at -60dbfs level by measuring the THD+N A-weighted. That plot shows the residual THD+N (A) of the LNA at 1-10mV level sweep. 2mV corresponds 2V 0dbfs DAC like CS43131, and -82db needs to subtract 60db to get AES17 DR = 142db(BTW, 10x better than APx555b). A typical high-performance DAC with 5Vrms 0dbfs could be measured down to 150db and so on.
APU_AES17DR.PNG



b) vinyl ripping with 34db of the gain for a MM cartridge, and 60db for the MC. Cosmos ADC has internal DSP where is implemented RIAA+subsonic filter.
Example:
Cosmos APU+Cosmos ADC+RIAA.
MM case, 5mV 10ohm source APU gain 34db Cosmos ADC 1.7V sensitivity, THD+N -89db or -97.7db(A). Due to H2 and H3 being very low, SNR is also 97db(A).
2022-05-31_23-12-09.jpg

MC case, 250uV 10ohm source APU gain 60db Cosmos ADC 1.7V sensitivity, THD+N -66db or -78db(A). Due to H2 and H3 being lower than the noise floor, SNR is also 78db(A).
2022-05-31_23-06-43.jpg



c) microphones usage.
Example:
The gain of Cosmos APU preamp is +60db(1000x times), hence the scale units are nVrms(A). After the APU unit turns On, you can see 5s of settling,
from 6s the input noise reaches 130nV, about 9s 48V power was turned On, and after 1s reached 142nS. Hence, APU's phantom power adds just 12nV to the preamp noise, that's a nearly ideal result as I think ;) The test was performed with a shorted preamp's inputs(In+ to In-) and biased to the GND by 2x6.8kOhm resistor as a dummy condenser mic.
LNA_48V_noise.PNG
 
Last edited:

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
“d) to use a Pro condenser mic with 48V phantom power with the Preamp of Cosmos APU.”

This interests me greatly.

A pristine mic pre with 48V for stereo recording using two low noise wide band studio mics.

So, I would need 2x Cosmos APU + 1x Cosmos ADC?
correct, because Cosmos APU is a single-channel mono device. As I know, one Chinese user uses that with iOS for video production, that's overcomplicated solution to me.
 
Last edited:

julian_hughes

Addicted to Fun and Learning
Joined
Aug 23, 2020
Messages
657
Likes
901
I also request somebody state in a few simple sentences what this does. It's not an auxiliary power unit, nor accelerated processing unit. Analog processing unit. A microphone amp, but more?
Thank you. I thought it was just me.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,409
Likes
2,773
Location
South of China, SHZ area, - Слава Україні
Guys, this is tech stuff, you don't need that if have no plan to measure distortions, noise, or a dynamic range at finer levels than $30K Audio Precision analyzers do. Also, this is the tool(preamp) for ripping vinyl by the most "digital" approach(I may miss something but Cosmos ADC+APU combo is the unique offer on the market today), or this preamp may work as a good studio mic preamp with built-in high-quality 48VDC phantom power.
That's it.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
Twin-T but an active one.
Actually I got it. I have had both, active and passive. Passive with -60dB at base frequency, the main disadvantage was the dependence of passband attenuation on source and load impedance. The active with -80dB at base frequency, as shown in post #3. Opamp selection is crucial (noise, distortion). This notch approach has pros and cons. After I bought your Cosmos ADC, I stopped using the 2T notch. I do not like the single frequency limitation, not mentioning the inevitable effect on phase of distortion component. Now I would only need the switchable divider/gain selector software driven. But frankly, it is just playing the game. All this is inaudible.
 
Last edited:

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
This was a measurement protocol of a power amp measured with a 2T -60dB passive notch. Please note the corrections for the notch response. The soundcard used was Creative USB X-Fi HD.

2T_protocol.png
 
  • Like
Reactions: trl
Top Bottom