• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DIYINHK DAC Measurements

meooms

Member
Joined
Mar 12, 2021
Messages
5
Likes
1
Great work Michael! Your ES9016 with the opa1612 is really performing very nicely.

As you know, I did switch to the DIYINHK ES9038 board, with the TP IVY-III IV stages. Though subjectively the noise floor was quite a bit lower, your measurements clearly show the importance of good implementation for this chip. So, though very happy with the sound of the ES9038 board, I also ordered the COSMOS the see how my setup performs objectively.
But more importantly, it will give me the opportunity to optimize the power wiring/connections (this had much effect in your measurements to my surprise) and to try out the ES9038 THD compensation. The board has various options for power, e.g. 2 separate AVCC supplies, separate power for the clock.

I agree that buying the OKTO DAC8 pro is the way to go, but - although a total noob - I really like to tinker with this stuff as a hobby.
When I have results I will post them here, but it will take some time.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
8,490
Likes
10,446
-No reason to go with the ES9038 over the ES9016, I am also using the "good" channels of the ES9038, expect inner channels to be worse
-ES9016 + Ian Canada I/V + OPA 1612 is the second best 8+ channel DAC I have measured so far, only surpassed by the Okto DAC8 pro
-Still think that once you add power supplies, I/V stages, op amps, deal with turn on / off pops and add an enclosure the cost is quite high and you are better off with a commercial offering like the MOTU Ultralite Mk5 or Okto DAC8 pro

Michael
Surprised how much better the ESS9016 is doing, really quite clean result. Thanks for measuring and sharing!
 
OP
M

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
1,286
Likes
1,541
Location
Detroit, MI
Great work Michael! Your ES9016 with the opa1612 is really performing very nicely.

As you know, I did switch to the DIYINHK ES9038 board, with the TP IVY-III IV stages. Though subjectively the noise floor was quite a bit lower, your measurements clearly show the importance of good implementation for this chip. So, though very happy with the sound of the ES9038 board, I also ordered the COSMOS the see how my setup performs objectively.
But more importantly, it will give me the opportunity to optimize the power wiring/connections (this had much effect in your measurements to my surprise) and to try out the ES9038 THD compensation. The board has various options for power, e.g. 2 separate AVCC supplies, separate power for the clock.

I agree that buying the OKTO DAC8 pro is the way to go, but - although a total noob - I really like to tinker with this stuff as a hobby.
When I have results I will post them here, but it will take some time.

Nice! It will be interesting to see how the IVY-III performs, it definitely seems like an external I/V offers the potential for improved performance.

So far power supply wiring has been one of the most important variables, I guess I didn't mention it my most recent posts but all of these most recent tests were done with extremely short wiring directly soldered on the DAC board side. I think the short wiring and because I am only driving 2 channels is why the performance was so good compared to previous measurements. I also plan on trying out a silent switcher + LT3045 regulator power supply to see if it offers the same performance (at least on the stock ES9016 board to which I have it directly soldered).

Michael
 
OP
M

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
1,286
Likes
1,541
Location
Detroit, MI
Have a few more measurements with the stock ES9016 board and the Cosmos ADC.

First I wanted to see how the SilentSwitcher + LT3045 power supply performed, as noted in post #15 and previously measured with the MOTU M4 in post #16. This configuration has a pin header installed on top of the DAC board to interface with an I2S DSP like a miniSHARC and also has a local 100 MHz Crystek 575X oscillator installed on the DAC board. I also wanted to see how the DAC performed with more than 2 channels driven.

ES9016, SilentSwitcher PS, stock I/V, NE5532, miniSHARC DSP, local oscillator, 2 V output, 3.5 V ADC input setting, 2 channels driven. THD = -111.9 dB, N = -112.8 dB, THD+N = -109.4 dB
es9016_minisharc_2vout_3.5vadc_stereo.png


ES9016, SilentSwitcher PS, stock I/V, NE5532, miniSHARC DSP, local oscillator, 2 V output, 3.5 V ADC input setting, 4 channels driven. THD = -103.0 dB, N = -110.1 dB, THD+N = -102.2 dB
es9016_minisharc_4ch_2vout_3.5vadc_stereo.png


ES9016, SilentSwitcher PS, stock I/V, NE5532, miniSHARC DSP, local oscillator, 2 V output, 3.5 V ADC input setting, 8 channels driven. THD = -91.5 dB, N = -108.0 dB, THD+N = -91.4 dB
es9016_minisharc_8ch_2vout_3.5vadc_stereo.png


So clearly things fall off quite a bit from a THD perspective with more channels driven. I also see the same high frequency noise rise that I saw in previous measurements but with the shorter wiring the noise level remains quite low.

To try and isolate the issue I switched back to my stock ES9016 with the DIYINHK XMOS and linear power supply. I played a 1 kHz 24 bit test tone from LMS / squeezelite and used CamillaDSP to route to more channels.

ES9016, linear PS, stock I/V, NE5532, XMOS / CamillaDSP, 2 V output, 3.5 V ADC input setting, 2 channels driven. THD = -109.2 dB, N = -112.1 dB, THD+N = -107.4 dB

es9016_stock_2ch_2vout_3.5vadc_stereo.png


ES9016, linear PS, stock I/V, NE5532, XMOS / CamillaDSP, 2 V output, 3.5 V ADC input setting, 2 channels driven. THD = -97.5 dB, N = -112.1 dB, THD+N = -97.4 dB
es9016_stock_4ch_2vout_3.5vadc_stereo.png


ES9016, linear PS, stock I/V, NE5532, XMOS / CamillaDSP, 2 V output, 3.5 V ADC input setting, 2 channels driven. THD = -88.4 dB, N = -112.1 dB, THD+N = -88.4 dB
es9016_stock_8ch_2vout_3.5vadc_stereo.png


THD really ramps up with the linear PS but get none of the high frequency noise rise of the SilentSwitcher.

I also checked my Ultralite Mk5 with all 10 channels driven and I get absolutely no change in THD or noise. Definitely another thing to consider when comparing these DIYINHK boards to commercial offerings.

Michael
 
Last edited:

mga2009

Member
Joined
Nov 21, 2019
Messages
94
Likes
43
Michael,

If you would want more than 10 analog outputs to use with CamillaDSP, is it possible to use 2x FR 18i20, the first one to use all 10 analog outputs and also use ADAT output to feed the second interface thru ADAT input, and use the second interface`s 10 analog outputs?

Is it possible to control the volume of all 10 analog outputs + ADAT output on the firsr interface? (I assume not, and the only way to control all volume would the the source volume).

I guess that an AVR with good preamp output (Denon X3500 - 3600- 3700) could feed 2x usb interfaces controlled by CamillaDSP and have a "cheap" 20 channel active crossover system (3x 3way for LCR, 4x independent SW, 2x 3way surrounds)... or I am just dreaming?
 
OP
M

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
1,286
Likes
1,541
Location
Detroit, MI
Michael,

If you would want more than 10 analog outputs to use with CamillaDSP, is it possible to use 2x FR 18i20, the first one to use all 10 analog outputs and also use ADAT output to feed the second interface thru ADAT input, and use the second interface`s 10 analog outputs?

Is it possible to control the volume of all 10 analog outputs + ADAT output on the firsr interface? (I assume not, and the only way to control all volume would the the source volume).

I guess that an AVR with good preamp output (Denon X3500 - 3600- 3700) could feed 2x usb interfaces controlled by CamillaDSP and have a "cheap" 20 channel active crossover system (3x 3way for LCR, 4x independent SW, 2x 3way surrounds)... or I am just dreaming?

You can definitely get 18 channels of analog output as you describe with 2x 18i20s. First one would be connected via USB for CamillaDSP and the second one would be clocked by the first interface and would be a "dumb" ADAT DA converter. Although you can definitely get 20 channels of analog output CamillaDSP would only be able to process 18 channels.

Another option to get 20 channels of analog output processed in CamillaDSP would be to create an ALSA aggregate device as described here -> https://github.com/HEnquist/camilladsp-config, this should work as you can clock the second interface from the first.

You are correct that the 18i20 volume knob will only control analog output channels, not ADAT output channels. However, if you use CamillaDSP for volume control you can have it act on all 10 analog output channels as well as the 8 ADAT output channels.

Definitely seems like a feasible option. You will want to keep the chunk size down as much as practicable with resampling disabled (should be feasible if you are using analog inputs where the same clock is used for input and output) for latency purposes. I use CamillaDSP on my A/V setup (just simple 2 channel from an AppleTV) and have no issues with latency.

Michael
 
Top Bottom