• 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.

Measurement and Review of FX-Audio DAC-X6 DAC (compared to Topping D30 and Fiio E10K)

Mad_Economist

Active Member
Joined
Nov 29, 2017
Messages
103
Likes
143
#21
I think that is a function of the higher noise level of R&S measurement gear he is using.
In theory, that R&S should be pretty low noise - about 2uV by the specs http://lcweb2.loc.gov/master/mbrs/recording_preservation/manuals/Rohde & Schwarz Audio Analyzer UPL (brochure).pdf which isn't too much higher than most AP's whose specs I've seen. Perhaps the THD+N plot is more flat in this case due to the dominant noise source being in the analog stage, meaning that relative digital level (and the DAC's nonlinearity) doesn't make an impact until high outputs?

Edit: Upon a moment's reflection, that would only make sense if the plot was referencing distortion products and noise to an absolute level, as Mindbomb says. If the THD+n is relative to the fundamental, all that would make sense would be that this is distortion, rather than noise, based, and the analyzer is simply low enough in noise (and the FX Audio high enough in distortion) that none of the distortion is hidden in the noise floor.

Just create a signature and put association/title/name of the company there as you see in mine.
Gotcha! That should be set up now. Just wanted to check, I know some places prefer to have members of trade or others with potential biases permanently flaired by admins.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#22
Ok, I think the reason I had trouble with that graph is that i didn't realize the units were dbv or some absolute type of unit, explaining why the noise would stay about the same even though signal was really low.
Ah, yes, that would explain the difference. :) Sometimes the simple answer evades me. :D
 
Joined
Nov 1, 2017
Messages
79
Likes
49
#23
Why does the headphone out of the E10K measure such high output impedance in your review? It's advertised to be 1 Ohm.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#24
Why does the headphone out of the E10K measure such high output impedance in your review? It's advertised to be 1 Ohm.
Let me double check my fixture to be sure it is reporting the right results.

I will be back! :)
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#25
OK, the fixture was fine but I had an operator error. :) When I saw the output oscillating, I lowered the volume with load, but forgot to go back and test the unloaded output at the same level. With that correction, the impedance is indeed a lot better:



I also added some filtering which made the results more reliable for the unloaded output. The impedance at 1 Khz is 1.4 ohms for Fiio and increases to 2 ohms at 100 Hz.

I also retested and updated the FX-Audio and my laptop and updated the OP. Thanks for prodding me to double check these results.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#27
Could you test the output with the gain jumpers (https://www.reddit.com/r/headphones/comments/5rk667/fxaudio_dac_x6_gain_jumpers/) set to "high"?

Additionally, would a loaded output vs. THD+N plot be possible? Thanks
Here you go. I modified the THD+N test to show the classical % value in log scale. Sampling rate is 48 Khz. There is some minor change in noise level based on sampling rate used.

This is for headphone output using 33 ohm load. I have also shown the unloaded output for the high-gain setting (it makes no difference in low gain mode).
Volume was set to max and digital input changed over coax input. Only one channel is loaded. I need to build by dummy load to do stereo testing.

headphone THD+N %.png


The unit ships with the performance shown in yellow. By moving two jumpers -- one for each channel -- you get the graphs in red/green (loaded/unloaded).

Given this data, if you listen in low gain setting at 70% or higher, the standard jumper setting works better with no clipping and similar to better signal to noise ratio. I think this is the most common range and as such the manufacturer has made the right choice to make this the default (and not document this jumper).

Setting the jumper to high gain will give a boost but you run into clipping at 2.14 volt and distortion skyrockets. As such its only use would be for lower volumes ironically where it provides better signal to noise ratio.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#28
While I was doing this testing, I also did a frequency response measurement on the headphone output:

headphone frequency response.png

You can see the much higher output levels with high-gain jumper setting.

Sadly we also see pretty significant roll off starting at just 5 Khz! We are taking 5 to 7 db drop by the time you get to 20 Khz. Pretty significant. In that sense it makes a good pairing with headphones that are too bright. I measured my Exasound E32 using the same setup an it is ruler flat while having more output. So some of the money you left in your pocket cost you the highs! :)
 

Mad_Economist

Active Member
Joined
Nov 29, 2017
Messages
103
Likes
143
#29
Setting the jumper to high gain will give a boost but you run into clipping at 2.14 volt
That is very odd. The output opamp in the DAC-X6 is meant to be the TPA6120A2, which shouldn't clip for such a low current output.

That, together with substantial high frequency rolloff (seriously, did someone fat finger a number while making the BOM? That can't have been intentional, surely) seems like a pretty poor choice. Very sad to hear.

Any chance you might be able to review the other of the cheap TPA6120A2 DAC-amps, the Monoprice Desktop Headphone Amplifier? https://www.monoprice.com/product?p_id=11567

Edit: Wait, it clips at 2.14Vpp? That's... very odd. I'm almost wondering if there might be some defect on your unit, as I would expect users who have swapped the jumpers to be reporting that - that voltage isn't unreasonable for many headphones at normal listening volumes, particularly for the less sensitive headphones which a lot of people are using the DAC-X6 for.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#30
First, please keep questioning me. :) These are new measurements and until I have a strong body of data, there could always be problems.

That said, here is the same setup this time including my Exasound E32 THD+N performance:

FX-Audio DAC-X6 headphone THD+N % compared to Exasound E32 DAC.png


We see that the Exasound onset of clipping occurs at 2.67 volts with the 33 ohm load. At that output voltage, if you look at the cursor 2 data, it has just 0.003% distortion while the loaded FX-Audio DAC-X6 has shot way up to 8.2% of distortion.

To confirm this finding subjectively, I listened to FX-Audio DAC-X6 using Sennheiser headphones (which has a far higher DC resistance of 300 ohms or 10X my 33 ohm dummy load). On bass heavy material, it is hard to listen to it at full volume. It is quite loud. But tolerating for a couple of seconds, it did seem to be more distorted than volume turned down a few notches. This maybe placebo due to the fact that I know it is "supposed to distort" :). But it is the best I can do. I need my hearing for future music enjoyment and can't donate it to science this way. :D
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#31
And here is the frequency response test run augmented with Exasound E32:

FX-Audio DAC-X6 headphone frequency response comparison.png


So again these are with the same setup so I think the fixture is correct and so is the data.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#32
Joined
Oct 10, 2017
Messages
160
Likes
74
#33
This is yet another THD+N measurement versus frequency in stereo (pairs of lines for each):

View attachment 9629

What the heck is the FX-Audio doing there??? Your guess is as good as mine.
The FX-Audio seems to be doing a classic rendition of high frequency nonlinear distortion. I see lot of this. A CCIF IM (19+20 KHz) test would seem to be in order. A pair of swept tones a constant 1 KHz apart starting at say 5+6 kHz could also be enlightening. I have a ESS 9018-based USB stick that does this. Again it is not uncommon. Subjectively, this sort of thing might cause audible problems with certain fairly atypical but natural sounds like my infamous "Keys Jangling" sample.
 

mindbomb

Active Member
Joined
Nov 11, 2017
Messages
284
Likes
146
#34
I suspect that high frequency problem might have something to do with how this dac handles that specific sample rate. I wonder if there would be any improvement with 96khz. Also, is it possible to get these graphs with vrms rather than vpp. I think that is more common when talking about voltage.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#35
I suspect that high frequency problem might have something to do with how this dac handles that specific sample rate. I wonder if there would be any improvement with 96khz. Also, is it possible to get these graphs with vrms rather than vpp. I think that is more common when talking about voltage.
Might be. I will test and report back. And yes, I can report on any unit. Which one specifically you want me to repost with Vrms?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#36

Mad_Economist

Active Member
Joined
Nov 29, 2017
Messages
103
Likes
143
#37
I went ahead and ordered that even though I am trying to not keep buying stuff until I get through my backlog of gear to review. :)
Fantastic to hear! I look forward to seeing the review of it - there really are not many budget DAC-amps which can drive less sensitive headphones to high volumes, particularly those which require a bit more current than average, and with the DAC-X6 apparently out of the running, that Monoprice is among the last under $100. Hopefully its results are a bit less discouraging.
 

sssn

Member
Joined
Nov 1, 2017
Messages
79
Likes
49
#38
Seriously, even a Fiio E10K drive headphones like the HD600/650 and 250 Ohm Beyerdynamics decently. Or just get a seperate DAC and hook an O2/Magni to it. Or use your onboard as DAC even, some are suprisingly good as pure DACs.

I really see no purpose to these underengineered, cheap power bricks (Monoprice Desktop amp, FX-Audio DAC-X6, SMSL whatever and so on).
 

mindbomb

Active Member
Joined
Nov 11, 2017
Messages
284
Likes
146
#39
Well, the sense I have from the headphone community is that people aren't even interested in using high gain on the e10k, let alone needing a more powerful amp. So even though the dac x6 can achieve a higher output voltage, I don't know if that translates into any real benefits.
 

mindbomb

Active Member
Joined
Nov 11, 2017
Messages
284
Likes
146
#40
Which one specifically you want me to repost with Vrms?
The thd+n 1% graphs, and these type of graphs moving forward. Am I projecting or is vrms more familiar to people? Can you add the fiio e10k in there as well or does this test only apply to devices with a line in?
 
Top Bottom