ASUS xonar U7 and W10 APO DSP measurements

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey all

Today I measured the ASUS xonar U7 USB soundcard with analog loopback headphone to line-in. So no samplerate conversion!
Was interested in distortion measurements when using the APO dsp program. Wondered if the free signalprosessing program distorted when running FIR filters generated by rePhase.
The PC is a Lenovo cobylake i5 so a little overkill.
Will maybe try a 100euro 7" Atom PC later.

The short answer. Suprising low harmonic distortion!

The last picture is just a extra measurement of a steep HP filter with a almost sqare 1kHz 20 dB boost. (Just some "difficult" eq:))

Used 32 bit wavefiles as FIR input
 

Attachments

  • rephase.jpg
    rephase.jpg
    350.9 KB · Views: 584
  • APO111.JPG
    APO111.JPG
    165.3 KB · Views: 578
  • and phase.jpg
    and phase.jpg
    279.8 KB · Views: 556
  • 50hz.jpg
    50hz.jpg
    255.7 KB · Views: 558
  • HP with squareboost1k.jpg
    HP with squareboost1k.jpg
    268.6 KB · Views: 193
  • lp.jpg
    lp.jpg
    237.5 KB · Views: 563
Last edited:
I had to turn off APO on input to make stable loopback measurements.
Enclose pictures of closeup phase and SPL with the ends of the spectrum. Be aware of the high resolution so it dont look good but it is good:)

Just a little very LF distortion

Tried also to set 196k samplerate on the U7. Still 48k JAVA samplerate.

Some HF attunation. I prefeer the no samplerateconversion
 

Attachments

  • 48kSF 0 to 30 000 196 U7SF.jpg
    48kSF 0 to 30 000 196 U7SF.jpg
    243.6 KB · Views: 109
  • 48kSF dist 0 to 30 000.jpg
    48kSF dist 0 to 30 000.jpg
    299.3 KB · Views: 109
  • 48kSF 0 to 30 000.jpg
    48kSF 0 to 30 000.jpg
    242.1 KB · Views: 107
I enclose the pictures from REW ASIO measurements. Then no APO of course.
48k, 96k and 192k.

For HF FIR filtering, I liked the 96k measurements best.
Not the VHF:) THD of the 192k and the phase and SPL has nice and easy roll off at the top. Gives some room to eq an octave away form FS for 20-20k systems.
 

Attachments

  • ASIO96k.jpg
    ASIO96k.jpg
    264 KB · Views: 77
  • ASIO192k.jpg
    ASIO192k.jpg
    254.9 KB · Views: 94
  • ASIO96k dist.jpg
    ASIO96k dist.jpg
    327.3 KB · Views: 85
  • ASIO192k dist.jpg
    ASIO192k dist.jpg
    340.1 KB · Views: 86
  • ASIO48k.jpg
    ASIO48k.jpg
    308.5 KB · Views: 86
Last edited:
Also wanted to se how samplerate conversion is handled by Win10.
That is if both analog in and local mediaplaying is intended.
All measurements REW 48k samplerate Java driver.
1. All 48k samplerate
2. 96k out. 48k in. So no samplerateconversion on input.
3. 96k in and out. SRC both in and out.

No change in distortion
Not bad, but some HF attunation with SRC.
 

Attachments

  • All48kFS.jpg
    All48kFS.jpg
    244.6 KB · Views: 81
  • 96kFSOutrest48k.jpg
    96kFSOutrest48k.jpg
    249.7 KB · Views: 71
  • 96kFSInandOutREWjava48k.jpg
    96kFSInandOutREWjava48k.jpg
    245.4 KB · Views: 61
I'm not exactly surprised that distortion is dominant 3rd towards the high frequencies. The opamp model they chose is a stupid low power part with like 300 µA of quiescent current but 200 mA of maximum output, so output stage bias is ridiculously low and crossover distortion would be expected to be non-negligible. (The really stupid part is that they also used it on the headphone out, rather than a dedicated headphone driver.) A bit of class A bias might really help things out, though it would involve a current source of very low saturation voltage.
 
Guess ASUS tried to save some PC battery time?

To me the distortion is small enough. (Lower than most tweeters and poweramps)
The soundcard has a lot of small flaws, but no big ones imho.

I really like the low noise and that APO kicks in very soon after you turn on the PC.
 
I use U7 and APO and i want to report that when 3 output jacks are used there is a noise on one jack present. But this only when 3 jacks are used. With 2 or 1, output is OK. I can live with that noise at the moment (also have a NAS that makes noise, and the PC..) but it is not acceptable if you have a silent room. Maybe is my USB port (laptop) .. ot maybe the soundcard does not have enough power to output clear sound on all 6 or 8 outputs.
 
I use U7 and APO and i want to report that when 3 output jacks are used there is a noise on one jack present. But this only when 3 jacks are used. With 2 or 1, output is OK. I can live with that noise at the moment (also have a NAS that makes noise, and the PC..) but it is not acceptable if you have a silent room. Maybe is my USB port (laptop) .. ot maybe the soundcard does not have enough power to output clear sound on all 6 or 8 outputs.

I've added a ADUM4160 USB isolator on my Terratec USB soundcard and all noises are gone :)
But my Creative Xfi USB interface don"t want to operate with it and i've got the same issue as you... there is no silent card at any price :wave2s:

So, i'm planning to use a balanced outputs USB interface, i've tested a Mackie pro soundcard with balanced amplifiers and it seems to be the only viable solution (except super expensive audio transformers that are more expensive than a high qualty pro multichanel soundcard)
 
Last edited:
They perhaps have to stay under 500mA of total power consumption, it's only 2.5W :2c:
Sure, the more power draw, the more fussy bus-powered devices tend to be. However, it's not like there'd be dozens of opamps in this thing. LME49726, I think that was it. Quiescent 0.7 mA per channel nominal, output current >300 mA. It looks very good on paper, but in reality things drastically worsen into low impedances and at high frequencies, so the high output current capability must be for driving phone/handset speakers or something, not hi-fi applications. Definitely not a good choice for a headphone driver. Performance is not looking great even with a 300 ohm HD600, let alone anything more demanding.

Looks like there is an U7 MkII model out now, I wonder what kind of changes it contains?
 
Sure, the more power draw, the more fussy bus-powered devices tend to be. However, it's not like there'd be dozens of opamps in this thing. LME49726, I think that was it. Quiescent 0.7 mA per channel nominal, output current >300 mA. It looks very good on paper, but in reality things drastically worsen into low impedances and at high frequencies, so the high output current capability must be for driving phone/handset speakers or something, not hi-fi applications. Definitely not a good choice for a headphone driver. Performance is not looking great even with a 300 ohm HD600, let alone anything more demanding.
Looks like there is an U7 MkII model out now, I wonder what kind of changes it contains?

USB powered interface are not a good choice to drive large headphones there is much better ones on the market designed for this purpose IMHO
Here is the ASUS offer for that purpose :
https://www.asus.com/fr/Essence-Hi-Fi-Audio/Essence_III/

Headphone listeners are taken in acount, and their dedicated products sound perfect. My Sound Blaster X-Fi Surround 5.1 Pro SBX is a real gem, but the annoying noises are a real pleasure killer, it's a real heartbrake and there is no mainstream consumer solution :t_ache:
 
I maybe get the time to see if there is any noise on mu U7 with > 2 outputs this weekend.
http://www.ti.com/lit/ds/symlink/lme49726.pdf
10k load is low distortion. Low impedance can be in the 0.1 % range.
What is the distortion in your headphones compared to the U7.
That is: Can anybody measure the headphone distortion and the U7 distortion at the same time. Please post results.
Theoretical statements is not very interesting to the thread starter:)
 
Got torture with Creative X-Fi 5.1 - very loud squeaky high-pitched noise.
Not sure if it is just poor design or defective piece, but the problem is on the 5V USB rail, caused by the soundcard itself. Made RCLC filter only to find out that 5V USB from the computer is actually cleaned FROM the soundcard, and not vice versa. Tried to filter the hell out of 5V and 3V3 for DAC and add better ground, but it´s still not so good I would want it. So thinking about Asus U7 or iConnectAudio2+ with ballanced outputs. Is any of that a good choice?
 
My use of the Asus U7 is FIR filtering for 2 x 4 way speaker.
Asus U7 8 ch into Denon surround receiver with analog 7.1 input.

See there is quite a lot of 50 Hz with harmonics noise. No signal out from measurement soundcard results in input at -85dB FS total. Dominated by 50 Hz and harmonics
Quess there is quite some groundloops with 8 ground wires from u7 to denon.
Then -85 dB is not that bad in an unbalanced system.
And I measured and sent signal with same soundcard so there is another source of 50Hz noise

I must connect the speakers later and check if it is annoying or just drowning in room noise.

Balanced output and balanced amps gives lower noise. 50 Hz noise is often all gone. But at a cost. This thread is for low cost Asus U7 with unbalanced out.
(It might be possible to make an balanced 4 output with some special cabels and inverting the neg output, but then there are a lot of nice bal out soundcards to buy instead.)
 
Last edited:
iConnectAudio2+ with ballanced outputs. Is any of that a good choice?
https://www.thomann.de/fr/iconnectivity_iconnectaudio2.htm?ref=search_rslt_+iConnectAudio2_388136_0
It seems to be a unknown gem ?
But there are only 4 outputs on the back, no more.
there are a lot of nice bal out soundcards to buy instead.
These ones ?
https://www.thomann.de/fr/tascam_16x08.htm
https://global.focusrite.com/usb-audio-interfaces/scarlett-6i6
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.