• 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.

Review and Measurements of HIFIMAN EF2A DAC and Headphone Amp

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
12,414
Likes
4,244
Location
Seattle Area
#1
This is a review of the HIFIMAN EF2A tube DAC and headphone amplifier. It is on a kind loan from a forum member. The retail price is $169 USD but I see it on Amazon for $139 plus shipping. I seem to recall it going on massdrop also. This is a small unit as you can see here, dwarfed not by one but two Audio Precision analyzers;

Hifiman EF2A DAC and Headphone Amplifier Review and Measureemnts.jpg


Typical of these "hybrid" headphone amplifiers, the tube is used as the input stage and the output drive is transistors. The see through lid shows a row of power transistors which hints at higher than normal output power.

I find the overall look cheesy and DIY type. The volume control though feels good. An external brick powers the unit.

Inputs are both digital in the form of USB and analog RCA. I tested the unit both ways.

Let's get into the measurements and see how she did. As usual, if you are not familiar with what these graphs mean, refer to my tutorial on understanding audio measurements: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/.

Measurements
Let's start with a test of jitter and noise:
Hifiman EF2A DAC and Headphone Amp Jitter Measurement.png


As we see, the performance of HIFIMAN EF2A is nothing to write home about. We see numerous distortion and jitter spikes together with elevated noise level (I matched the levels to Fiio Q1).

Turns out there are much more serious problems here. Check out this time domain output of a 1 kHz tone at 0 dbFS through USB input:
Hifiman EF2A DAC and Headphone Amp 1 kHz Distortion Measurement.png


Yes, the output is clipped. And no, this is NOT, let me repeat, NOT caused by the output stage of the amplifier. The output is clipped regardless of volume control position. The USB DAC is overdriving the input stage of the amplifier causing the clipping. I had to reduce the output by 10+ dB to get rid of most of it.

Same tests through analog input showed clean output so clearly this is a design problem with the USB DAC implementation.

Given the bad state of affairs on USB input, I tested the output power/voltage versus distortion using the analog input at different loads:

Hifiman EF2A DAC and Headphone Amp THD vs Power Measurement.png


Distortion+noise curves are rather high even here although within the manufacturer specs. On the positive front, even at 33 ohm load, the output can go as high as 5 volts. At the kneed of clipping point, it was about 3 volts resulting in 0.29 watts.

More bad news awaited us when measuring the output impedance:

Hifiman EF2A DAC and Headphone Amp output impedance Measurement.png


As you see, at 31 ohms, it is worse than anything I have ever measured by far. This means the output of the unit will vary substantially with the headphone load you put on it.

Not shown is frequency response which was pretty much flat up to 40+ kHz (using analog input).

Given the above issues, and in the interest of getting this review out, I did not perform more tests. I think you get the picture here.

Subjective Listening Tests
We often argue if distortions are audible. Well, this is a case where this is definitely true and not true! It is true in that my son and I heard the problem with USB clipping first in listening test. I was getting ready to do an AB test and to match levels, I played a 1 kHz tone, and to our amazement, it sounded so different on the HIFIMAN EF2A compared the FIIO Q1. The pitch was completely different. As my son nicely put it, "if you can tell they sound different with the headphone on your lap, there is a definitely problem here!"

I then set up a comparison between the EF2A and my RME ADI-2 Pro. Surprisingly, it was difficult to hear much if any difference on actual music. The reason is that my reference tracks are the ones I enjoy playing and do not have high amplitude. My son suggested that K-pop music may show the clipping problem better. So we hunted some on Tidal and there, there was some exaggeration of high-frequencies in the EF2A.

On the positive front, the EF2A could keep up with the ADI-2 Pro with driving even my power hungry Sennheiser HD-650 with authority. There was plenty of power there to drive it to loud listening levels.

Conclusions
As a USB DAC+headphone amplifier, the HIFIMAN EF2A is clearly a broken design. The DAC stage is overdriving the amplifier. There is no excuse for this, nor does it have any positive attribute. Its output impedance is also exceedingly high. For these reason, I can't recommend this unit.

That aside, or if you have an external DAC, the unit performs fine. There are no euphonic "tube sound" that either myself or my son could detect so don't buy it because you think the tubes are doing something for you. They are not.

As always, questions, comments, corrections, etc. are all welcome.

Edit: hardware teardown and pictures here: https://www.audiosciencereview.com/...iman-ef2a-dac-and-headphone-preamplifer.3462/

Edit 2: And horrible linearity results in this post: https://audiosciencereview.com/foru...an-ef2a-dac-and-headphone-amp.3402/post-86238
----
If you like this review, please consider donating funds for these types of hardware purchases using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).

P.S. Other non-DAC reviews are coming. I promise. :) It is just that I have a lot of loaned DACs that I am having to review so that I can return them.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
12,414
Likes
4,244
Location
Seattle Area
#2
Oh, I forgot to mention: in listening tests, changing headphones vastly changed the volumes with the HIFIMAN EF2A. Its high output impedance causes that and as a result, your subjective listening experience may be very different than mine.
 

Elix

Member
Joined
May 15, 2018
Messages
15
Likes
4
#3
Thanks! I liked your distortion+curves graph with different loads. It will be most welcome in your other tests to come.
Amir, is your TODO thread correct in depicting the order in which you're planning to do reviews? I don't see Sabaj DA3/Topping NX4 DSD which we were talking about in the other thread and you said that you have them. :(
 
Joined
Feb 22, 2018
Messages
31
Likes
8
#4
Would you mind doing a channel balance vs gain measurement like in your last review?
This seems like the type of product that could become a reference for its negative traits.
 

Valorum

Member
Patreon Donor
Joined
Jan 16, 2018
Messages
21
Likes
8
#5
Awesome, thanks for reviewing my unit, Amir!

I did not want to divulge my impressions prior to your testing and potentially color the review. This was my first DAC/amp in any capacity, so I did not have a frame of reference or expectations.

Issues I noted:
  • Extremely high gain, nearly unusable with low impedance headphones, even with the HE-400i (which is the advertised pairing on their website), to the point where I would have to decrease output on the source to maybe 50% to get any usability out of the gain knob
  • Toward the bottom of the adjustment range of the gain knob, there's noticeable channel imbalance
  • I noted static or other artifacts during playback at low volume (like doing homework), and thought it was my headphones initially. I discovered I liked just listening to them straight from the Macbook Pro's output more than through this unit
  • An obnoxious *BWAMP* noise/pop when turning the unit on with headphones plugged in. It was loud enough I was concerned about damaging the drivers
I was curious to see if this unit performed more poorly than the Schiit products.

Amir, any idea how many bits of resolution this unit can deliver?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
12,414
Likes
4,244
Location
Seattle Area
#7
Amir, is your TODO thread correct in depicting the order in which you're planning to do reviews? I don't see Sabaj DA3/Topping NX4 DSD which we were talking about in the other thread and you said that you have them. :(
The list is more or less correct of the gear I own. It is usually out of date on loaned equipment.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
12,414
Likes
4,244
Location
Seattle Area
#8
Amir, any idea how many bits of resolution this unit can deliver?
I chose to not run the linearity test due to the clipping over USB. When I get a chance I will give it a shot and see how it does.
 

gvl

Active Member
Joined
Mar 16, 2018
Messages
258
Likes
52
Location
SoCal
#9
That aside, or if you have an external DAC, the unit performs fine. There are no euphonic "tube sound" that either myself or my son could detect so don't buy it because you think the tubes are doing something for you. They are not.
I can't think of any useful function of tubes apart from coloring the sound, if there is none I don't see why you need tubes at all. Also, does this unit have stock tubes? Just thinking that clipping can be due to non-original out of spec tubes being used, unlikely but just a thought.
 

Valorum

Member
Patreon Donor
Joined
Jan 16, 2018
Messages
21
Likes
8
#10
I can't think of any useful function of tubes apart from coloring the sound, if there is none I don't see why you need tubes at all. Also, does this unit have stock tubes? Just thinking that clipping can be due to non-original out of spec tubes being used, unlikely but just a thought.
The unit does not have stock tubes. It has aftermarket tubes, Mullard/RTC 5654 (6AK5W), that were included when I bought it.

According to one Amazon review, those are "the best."
 

gvl

Active Member
Joined
Mar 16, 2018
Messages
258
Likes
52
Location
SoCal
#12
The unit does not have stock tubes. It has aftermarket tubes, Mullard/RTC 5654 (6AK5W), that were included when I bought it.

According to one Amazon review, those are "the best."
Then they must be, huh? Are they NOS or modern production? How many hours on the tubes?
 

Valorum

Member
Patreon Donor
Joined
Jan 16, 2018
Messages
21
Likes
8
#13
Then they must be, huh? Are they NOS or modern production? How many hours on the tubes?
I must admit I have no idea. I've not played with tubes before, and the previous owner of the unit installed them and used them before I did. I have maybe 10-20 hours on them, I'd guess.
 

gvl

Active Member
Joined
Mar 16, 2018
Messages
258
Likes
52
Location
SoCal
#14
I must admit I have no idea. I've not played with tubes before, and the previous owner of the unit installed them and used them before I did. I have maybe 10-20 hours on them, I'd guess.
It would be interesting to try another tube set, different type. There seem to be an infinite NOS supply of inexpensive military spec GE JAN 5654W on e-bay. I tried them in my mini SET tube amp, they are decent sounding, not much coloration, maybe a bit dry.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
5,114
Likes
1,158
Location
Riverview, Fl
#15
Also, does this unit have stock tubes? Just thinking that clipping can be due to non-original out of spec tubes being used, unlikely but just a thought.
What would so squarely limit the output of a similar replacement tube vs the specifed original?

(I'm tube ignorant here for something like this. Seems like it would work, at least not clip so cleanly)
 

Valorum

Member
Patreon Donor
Joined
Jan 16, 2018
Messages
21
Likes
8
#16
What would so squarely limit the output of a similar replacement tube vs the specifed original?

(Tube ignorant here for something like this. Seems like it would work, at least not clip so cleanly)
I will not pretend to have any understanding of how tubes work their black magic, but could the non-stock tubes amplify the signal more than intended until it runs into a ceiling imposed by other components?
 

gvl

Active Member
Joined
Mar 16, 2018
Messages
258
Likes
52
Location
SoCal
#17
Not an expert either but triode tubes are just like transistors, if they fully "open" (or "close" if you will, meaning max current flow) too early they will clip. There maybe some variance between tube models in their specs, and if the input circuitry supplies too high of a voltage on the grid of some peculiar tube this is what happens.
 

Valorum

Member
Patreon Donor
Joined
Jan 16, 2018
Messages
21
Likes
8
#18
It would be interesting to try another tube set, different type. There seem to be an infinite NOS supply of inexpensive military spec GE JAN 5654W on e-bay. I tried them in my mini SET tube amp, they are decent sounding, not much coloration, maybe a bit dry.
If I liked the rest of the unit, I might consider trying other tubes. I found it rather uninspiring on the whole.

I would be curious as well to see how different tubes impacted the testing.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
12,414
Likes
4,244
Location
Seattle Area
#20
By the way, I love the way you annotate/comment on your graphs. It makes it so much easier/faster to understand.
That's great feedback. :)
 
Top Bottom