Meizu Hifi USB-C Headphone Adapter Review

[amirm]

This is a review and detailed measurements of the Meizu Hifi USB-C, smartphone DAC and Headphone adapter. I can't find a price on it but seems to be listed for US $35. it is on kind loan from an overseas member.

You can't see it in this picture but there is a cool circle showing what is inside of it:

As long as you have a headphone plugged into it, Windows 10 automatically installs the driver for it. And works down to 24 bits with no issue.

My new custom PC has a USB-C native port on it and that is how I tested. But I also tested it with the USB hub in my monitor which I used to use for much of my DAC testing. See a comparison of that below.

Headphone DAC & Amp Audio Measurements
As usual we start with our dashboard view:

There is a lot of good news here starting with 2 volt output instead of the typical 1 volt and very low distortion. Not so good news is that noise level must be high to cause THD+N/SINAD to be much worse than the distortion alone indicates. We are seeing much worse measurement here than the spec as indicated.

A test of dynamic range shows the problem:

We are 30 dB shy of the stated dynamic range! Yet I have seen measurements that show 120 dB dynamic range. So I changed USB ports to that of my monitor and got this:

Yes, it got much worse! Filtering of USB noise seems to be absent causing such variations. So who knows what performance you get from your device. As it is, the Meizu Hifi lands near the bottom when it comes to driving sensitive IEMs with least amount of noise:

Jitter test shows abundance of interference likely coming from the PC over USB bus:

Filter them out and test linearity and results are excellent!

Sad.

Headphone Amp Power Measurements
Power is good into both 300 and 33 ohm loads:

Edit: forgot to post the output impedance measurements:

Conclusions
The heart of the Meizu Hifi is good but implementation not. I suggest getting the Meizu Hifi Pro which performed a lot better and not much more money.

------------
As always, questions, comments, recommendations, etc. are welcome.

I need to buy winter clothing for the panthers so that they come and pose for picture on speakers in our cold garage. Stop thinking for yourself and donate to keep panthers warm using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

 

[oliveoilthansnakeoil]

Hm... Hidizs s8(or tempotec sonata pro), meizu non pro and pro...
It seems they all have high distortion as an amplifier.

 

[IVX]

I have the same dongle on my desk, it is much better in fact. I using Redmu K20 Pro as a USB host streamer. SNR -130.5db(A)(i.e. less than APx555 able to measure!), no any bushes at 10kHz. Maybe a bad sample?

 

[amirm]

I have the same dongle on my desk, it is much better in fact. I using Redmu K20 Pro as a USB host streamer. SNR -130.5db(A)(i.e. less than APx555 able to measure!), no any bushes at 10kHz. Maybe a bad sample?

I don't think it is a bad sample. I think it picks up noise from USB so what you measure highly depends on how you connect to it. Your adapter likely is much cleaner than any PC USB.

 

[vitalii427]

I don't think it is a bad sample. I think it picks up noise from USB so what you measure highly depends on how you connect to it. Your adapter likely is much cleaner than any PC USB.

JFYI Redmi K20 Pro is a smartphone from Xiaomi.

There was a lot of controversy about this dongle on other threads. Amir, could you please run any quick test with smartphone to check noise performance?

 

[amirm]

Amir, could you please run any quick test with smartphone to check noise performance?

Here it is when played from Roon on my phone. No idea what pipeline is being used (I set it to 24 bits in Roon):

 

[IVX]

amirm
my dongle is prepared for the tough real-world USB 5V rail and contains a few quite expensive 1000uF 6.3V 2mm profile tantalum caps($1.5/pcs). BTW, 5 years ago such caps didn't exist. But people gonna use Meizu mostly with a mobile phone, either a battery-powered laptop, so would be reasonable to measure with such type of streaming source.

 

[IVX]

amirm
please try to add AUX-0025 to cut USB HF noise, I didn't mention but I always use that filter. Without it AP gain may don't go high enough.

 

[vitalii427]

amirm
my dongle is prepared for the tough real-world USB 5V rail and contains a few quite expensive 1000uF 6.3V 2mm profile tantalum caps($1.5/pcs). BTW, 5 years ago such caps didn't exist. But people gonna use Meizu mostly with a mobile phone, either a battery-powered laptop, so would be reasonable to measure with such type of streaming source.

Is this only about 9038D or 9039S too?

 

[bobbooo]

Here it is when played from Roon on my phone. No idea what pipeline is being used (I set it to 24 bits in Roon):

View attachment 46313

Could you run this test again using Neutron Player (free trial here) or USB Audio Player Pro? As unlike most apps, they are definitely able to bypass the Android audio mixer, which often degrades audio.

 

[IVX]

vitalii427, yes sure. First 9038S had 3000uF, next G2 2000uF.
bobbooo, it is not about that because the noise floor is Ok and landed at -150db, I think it is about physical HF USB noise.

 

[bobbooo]

bobbooo, it is not about that because the noise floor is Ok and landed at -150db, I think it is about physical HF USB noise.

Yes but the measurements show even more noise with the phone than PC, which might be explained by the Android audio mixer. It would be useful if Amir could get proof of this by using Neutron Player or UAPP, which are known to bypass the Android mixer.

 

[Johnb]

I have the same dongle on my desk, it is much better in fact. I using Redmu K20 Pro as a USB host streamer. SNR -130.5db(A)(i.e. less than APx555 able to measure!), no any bushes at 10kHz. Maybe a bad sample?

The Pro was also disappointing when measured by Amir, -120 THD, yet only Sinad around 100db. I was hoping that the test of the non-pro model would show some setting had been wrong or driver issue with the original review of the Pro, but both are performing similarly with regard to noise. I know that you, IVX, keep getting much better numbers for noise, so I hope Amir might get to the heart of the discrepancy. The check with smartphone usb is what immediately came to my mind as well, but no good.

I bought this same dongle (non-pro) in hopes that noise would be minimal loaded only by my portable amp (Topping NX3S). Alas the S8 and variant would have been a much better choice!

 

[IVX]

Johnb
non-pro it is bare Cirrus DAC CS43131 and it makes what was promised in its datasheet. Of course, Meizu(actually, any other tiny dongle model) didn't find extra room on the PCB for any attempts to make USB power cleaner, and now you are hostage of USB port quality.

 

[Luyil]

Hi @amirm , Did you get the output impedance of it? Thanks

 

[amirm]

Hi @amirm , Did you get the output impedance of it? Thanks

Yes. It is very good. Will post tomorrow.

 

[CopperFox]

It would be interesting to see how something like those independent USB power purifiers that were reviewed a while ago would affect the results here. Although they were in the $100 price range, there are some cheaper such products as well. Fiio released their "FiiO LA-UA1 USB Power Purifier" ($25) just this month. If something like that would work well with an usb dongle, the combined price would still be low. Perhaps the Meizu dongle or some equivalent could be used to test such products in the future?

Edit: the Fiio purifier does not support a simultaneous data connection though. Just power.

 

[Veri]

It would be interesting to see how something like those independent USB power purifiers that were reviewed a while ago would affect the results here. Although they were in the $100 price range, there are some cheaper such products as well. Fiio released their "FiiO LA-UA1 USB Power Purifier" ($25) just this month. If something like that would work well with an usb dongle, the combined price would still be low. Perhaps the Meizu dongle or some equivalent could be used to test such products in the future?

I've sent @amirm a Singxer UIP-1, hope he can try it with this dongle since it seems an ideal candidate for extra filtering.

 

[confucius_zero]

I guess that settles it... Pro dac wins, obviously.

 

[IVX]

I measured the output Z of this DAC by Agilent impedance-meter at 1kHz, without cable, on the plugged 3.5mm jack directly, there is around 300mohm.