• 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). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

ASUS Essence One mods

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#1
Hello,

I was able to hear some hum noise...a mixture between 50Hz and 150Hz, especially when using my sensitive AKG K550 cans with volume to the max, of course...no input source connected! I know these sensitive 32-Ohms cans can be driven by only few dozen of mV only instead of 7V when volume to the max, but I know this DAC/headamp combo measured some hum by Archimago (here and here) and also by other people's RMAA, so I needed a way to test it myself without having a decent ADC around. So yes, the hum is there and audible.

If I was reclining or moving away the toroid trafo then hum was gone, actually...if I was opening the case the hum os not there anymore, so most likely the thick metallic case bounced back the EMI field into the nearby potentiometer or something. So yes: case open = no more hum; case closed and toroid in a different angle = hum lowered a lot...quite strange.

However, I fixed the hum noise for good by wrapping twice a band of "grain-oriented silicon steel" (G.O.S.S. band) around the toroidal transformer. Now there's absolutely no audible hum at all not with AKG K550 anymore and not even with my very sensitive 16-Ohms Superlux IEMs (quiet night, mute=ON, pot. to the max. = no hum at all!).

The G.O.S.S. band I got it from a friend who works as engineer for Petra-toroid (I got it for free, they don't sell this). Also Talema is implementing this as well in Hi-Fi equipment: http://www.nuvotem.com/en/products/gossband_large.shtml. So yes Sir, Faraday knew some shit. :)

E1_hum_01.jpg E1_hum_02.jpg E1_hum_03.jpg
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
19,648
Likes
20,902
Location
Seattle Area
#2
So yes: case open = no more hum; case closed and toroid in a different angle = hum lowered a lot...quite strange.
Quite strange indeed!

BTW, I have noticed that with my 300 ohm headphones, the leads pick up fair amount of measurable noise! Impedance is high enough as to allow that. Don't see that with < 100 ohm headphones.
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#3
EMI field is trapped inside the thick case and bounces back into the PCB traces and, in Essence One's case it bounces back into the volume pot.

L.E.: Hope you do notice the metallic band strapped around the toroid that helps in lowering the 50Hz hum to inaudible.
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#4
Here's a very "not so nice" ground loop between the -12V power rail of the PSU and the PCB itself and how this looks on a Tek and on a Pico:

E1_-12V_02.png E1_-12V_ground_loop_03.jpg

Here're 2 ways to resolve this (directly on the plug or soldering on the backside of PCB):

E1_-12V_ground-loop_02.jpg E1_-12V_ground-loop.jpg

Now -12V looks fine:
E1_-12V_01.jpg

L.E.: A simple wire that connects the -12V ground to the PCB main ground resolves this. I wonder how could ASUS missed a ground connection inside the wires between the PSU and the PCB.

Also, here's how -12V looks like at the PSU side and not to the PCB side: E1_+12V_at_PSU_plug.png
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#5
Speaking about the poor grounding and ground loops, here's the +12V:
E1_+12V_01.jpg

And here's after resolving some grounding issues on the PSU's PCB:
E1_+12V_02.jpg

Issues I was speaking about:
E1_PSU_mod_03.jpg - a wire between the power regulator and the ground (too thin ground traces and probably no dedicated layer for ground plane)
E1_PSU_mod_02.jpg - 2 more thick wires to improve the grounding for +5V, +12V and -12V rails.
E1_PSU_mod_01.jpg - increased capacitance on all 3 rails
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#6
The main issue was with the +5V rail, because this DAC/headamp combo device was using only 1 x 5V power source for everything: relayes, the ARM chip inside, the DSP chip and for both PCM1795 DAC. Quite lame, isn't it? And some PCB traces were having somewhere around 30cm length, especially the ones from the PSU till the PCM1795's analog Vref supply pins. :(

E1_+5V_01.jpg - original +5V PSU on a Tek
E1_+5V_02.jpg - after replacing 2940 regulator with a NJM7805 it looks a bit better, but still nasty
E1_+5V_03.jpg - after improving PSU's grounding as on the above post + NJM7805 + CRC filter (1800uF+3.3ohms+2200uF) + 2x10uF tantalums & 1x220uF SILMIC2 the +5V looks better indeed
E1_LT1963A.jpg - changing the regulator with a LT1963A
E1_+5V_04.jpg - after all of the above steps the +5V looks satisfactory at the PSU's plug side
E1_+5V_on_DAC_pins.jpg - although, measured on the Vref pins of both PCM1795 DACs I don't like the noise (thanks to the high-length PCB traces and the ARM chip nearby)
E1_cut_traces_01.jpg E1_cut_traces_02.jpg - decided to cut the +5V traces in the proximity of the two PCM1795 chips and to create a new +5V dedicated PSU to both DAC chips (I choose a NJM7805 having a max. of 40uV of noise) powered from the same transformer, just in front of existing +5V (before the existing +5V regulator)
E1_new_+5V_02.jpg - now I can sleep well, as this is what I measured at the Vref pins of the DAC chips.

Below is the cable coming from the DIY 2'nd +5V PSU:
E1_new_+5V_01.jpg E1_+5V_PSU_05.jpg

Despite of the dual PCM1795 DACs and good components around them, the lack of correct grounding techniques, ground-loops, poor PCB with very long power traces were causing the device to not measure very well (though Stereophile gave it around 19 bits or resolution) and to sound a bit grainy. Now, after all of the above mods, the Essence One DAC/amp sounds very well to my ears.

P.S.: A second identical combo DAC was measured and I got about the same measurement results with same Tektronic scope, so it was not an issue with my combo DAC only, it was probably a general issue.
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#7
If the above mods did really improved the output sound, now some easier mods that might or might not improve the sound:
E1_caps_out.jpg - paralleling the existing polarised Nichicon caps on the bottom of PCB with non-polarised WIMA's (couldn't notice any sound improvement on XLR/RCA outputs).

WP_20180401_21_52_17_Pro.jpg - 10-Ohms output resistors added by ASUS but not recommended by TI on the LME49860 datasheet. With 30-Ohms resistive load I was able to measure about 11-Ohms output impedance. I can't listen very well to 16-Ohms IEMs nor to 50-Ohms planar Fostex T50RP-mk3 because lower bass is somehow loosing...not thick enough and losing the intensity. However, with my other cans sound is perfect (AKG K701, Beyers DT880 600-Ohms etc.)

10ohms_resistors_mod.png - paralleling additional 10-Ohms resistors on existing 10-Ohms resistors from the PCB (thought I can just short-circuit these resistors anyway) lowered the amplifier's output impedance (when 1kHz signal applied and 30-Ohms resistive load is used) down to 2.9-Ohms. 16-Ohms IEM's and Fostex planars sound perfect now...couldn't believe what I was hearing and I was also doing A/B'ing with an external amp, so it's not my imagination.

With 15-Ohms load resistor sine waves look perfect and having the same amplitude across audio spectrum, so the response would be perfectly flat.

Screen Shot 2018-04-09 at 21.55.25.png - a 20Hz square-wave played shows no bass roll-off @ 20Hz.
Screen Shot 2018-04-09 at 21.55.44.png - 1kHz square-signal, although not representative
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#8
Forgot to say that in I/V stage the 4xMUSES01 add some more details and warm in the sound that makes it more pleasant.
I'm also running 2xMUSES02 in the LowPassFilter, but I'm unable to differentiate these opamps out of MUSES8820 for example.

E1_EMI_filter.jpg - I've also added a common-mode EMI filter on the 230V mains cables and a bit of ground decoupling with a cap and a 5-Ohms/2W resistor (I've seen this decoupling technique inside Meier Corda's Jazz, I think). This will be decreasing the hum in case of a ground-loop created with the XLR/RCA interconnects and an external amplifier.
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#9
Quite strange indeed!

BTW, I have noticed that with my 300 ohm headphones, the leads pick up fair amount of measurable noise! Impedance is high enough as to allow that. Don't see that with < 100 ohm headphones.
If cables are long enough then I believe it's noise captured from the headphones cable, right? It's probably outside audible range, right?
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#10

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
2,085
Likes
3,089
Location
Gold Coast, Queensland, Australia
#11
Well, I admire your tenacity in trying to re-engineer what was clearly yet another badly designed product. :)

Hum and buzz from an unshielded toroid at close proximity is just plain laziness and cost cutting. Mu-metal/grain oriented steel have been used for many decades around toroids and to prevent EI transformers from radiating crap into associated circuitry.

I don't think a single piece of my older HiFi gear with toroidal transformers in it isn't either potted and shielded or with EIs, not having a mu-metal flux strap someplace, along with careful mounting.
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#12
Worth adding https://www.audiosciencereview.com/...urson-sparkos-sonic994-1612.3305/#post-153240 here, because is somehow related. Seems that default NE5532P easily outperforms the MUSES01 when used in I/V stage.

Not the best performer for sure, but it has an incredible low-noise, completely inaudible with my sensitive IEMs (SPL of 103dB/mw, 120.5dB/V @1KHz, 16 Ohms). On the headphones-output it has 2xLME49600 in the loop of 2xLME49720 (DC-servo, no caps on headphones-out).


NE5532P in I/V, MUSES02 in LPF (headphones-output jack)
Worth mentioning that adding Wyrd makes no difference in any of the spikes or noise-floor. Also, COAX input makes an identical graph (ASUS U7 used as transport).
 

JohnYang1997

Addicted to Fun and Learning
Joined
Dec 28, 2018
Messages
927
Likes
385
#13
You can try njm2068 to replace thr 49720 around the 49600. Some how njm2068 and opa2156 outperforms lme49720 and opa1612 in this application.
Also performance drops under load with output resistor over 0.1ohm. You may parallel 0.1ohm resistors there. 0 ohm resistor should works too.
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#14
Thank you, I will replace LME49720 with NJM2068 on one channel only and measure the THD+N to compare it.

If you're referring to the output resistors after the LME49600, I honestly don't care much, because I had it tested already with 16 Ohms IEMs and also with dummy 10 Ohms resistors and I got perfect neutrality and linear THD+N over the entire audio spectrum. Of course, if I'll ever encounter an issue in regard with the damping factor or if I suspect any oscillations, I will definitely remove the output resistors and replace them with inductors in parallel with resistors (less than 0.1 Ohms resistance), like here: http://www.twistedpearaudio.com/docs/linestages/ventus_ez_schematic.pdf.
 

spede

New Member
Joined
Jun 27, 2019
Messages
1
Likes
0
#15
How good is this amp/dac today? Essence One Muses unmodded compared to lets say Khadas tone board and JDS Atom combo.
Tried ifi micro se but it sounded worse than Essence headphone amplifier.
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
615
Likes
396
Location
Iasi, RO
#16
My measurements are all @ 1V RMS, but when doing those tests on 2V or perhaps 5V, usually THD+N gets better. Although, https://www.stereophile.com/content...-processor150headphone-amplifier-measurements gave 19 bits of resolution on 600 Ohms cans to the MUSES version.

Below measurements done with 1V RMS on headphones output, NE5532 vs. MUSES (it's on my modded Essence One combo):
ASUS_E1_1KHz-48KHz_Headphones_NE5532_IV_LME49720_LPF.png

NE5532 opamps in I/V and LME49720 in low-pass filter
-93dB of distortion means a SINAD of 15.5 bits


ASUS_E1_1KHz-48KHz_Headphones_MUSES01_IV_MUSES02_LPF (1).png

MUSES01 opamps in I/V and MUSES02 in low-pass filter
-83.5dB of distortion means a SINAD of 14 bits


Worth mentioning that in A/B tests 10 out of 10 tests were favorable to MUSES, I've no idea why, but with NE5532 it definitely measures a bit better. I'll try to find an original Essence One and test them both side by side.

Given the low price of the Khadas and Atom, I think it could be a better match.

P.S.: What I like most to my modded E1 MKii is the very low background noise and it's ability to drive cans up to 7V RMS. Of course, the built-in pre-amplifier is perfect for me to drive my Mackie monitors + subwoofer on RCA-out and my Matrix M-Stage HPA-3B on the XLR plugs.
 
Top Bottom