• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Incredible micro DACs

  • Thread starter Deleted member 65547
  • Start date
D

Deleted member 65547

Guest
This week I satisfied my curiosity and ordered a pair of micro DACs from Amazon, both in the 32/384 version but from different manufacturers. The price was roughly equivalent to a pizza and a beer each.

They arrived yesterday morning, with the white one impressively packaged, complete with a small metal box. The black one, more straightforward, arrives in a cardboard box but has the great advantage of allowing phone charging through a Power Bank while listening to music (quite useful!).

I believe the chips used are the same, CS46L41 + SA9302L, or at least that's what the white one declares, as I don't notice any differences in sound.

Well, they cost about the same as a pizza and a beer, so I expected performance to match... but no, they sound surprisingly good!

Unfortunately, the Amazon Music Unlimited app can't push Android beyond the limits set by Google, so it's limited to 24/48 (which is still much better than 16/44).

Switching to USB Audio Player Pro for other music, thanks to its custom drivers that happily bypass Google's music restrictions.

I won't claim that Mother Earth Jazz 24/192 sounds sound like the pair of Sanskrit MK3s that drive a class A amp in my stereo system, but they sound good.
Too good.
I suspect there's alien technology inside.



IMG_20231103_232831.jpg


Screenshot_2023-11-03-21-58-57-395_com.amazon.mp3-edit.jpg


Screenshot_2023-11-03-21-33-00-198_com.extreamsd.usbaudioplayerpro-edit.jpg
 
Last edited by a moderator:

IAtaman

Major Contributor
Forum Donor
Joined
Mar 29, 2021
Messages
2,409
Likes
4,165
If they can not do 384kHz it is waste of money. Sample rate it can handle is the most important aspect of any DAC.
 
OP
D

Deleted member 65547

Guest
If they can not do 384kHz it is waste of money. Sample rate it can handle is the most important aspect of any DAC.
My ears say that the biggest difference is made by 24 bits, even a modest 24/48 has a superior performance to a simple 16/44 CD quality.
However they reach 32/384, but I don't have a track suitable for the test.
 

DVDdoug

Major Contributor
Joined
May 27, 2021
Messages
3,035
Likes
4,002
This week I satisfied my curiosity and ordered a pair of micro DACs from Amazon, both in the 32/384 version
Any measurements? Manufacturer's aren't always honest.

Any cheap DAC/soundcard can play a 32/384kHz file because the drivers will down-sample to match the hardware limitations. It's just like printing a high-resolution photo on a low-resolution printer. The manufacturer can semi-legitimately claim that it "supports" 32/384.

There are very few 32-bit DACs and most 24-bit DACs are only accurate to around 20-bits so you're never really going to get 32 useful bits (or need 32-bits).

I'm not sure what my upper limit is but I can't hear to 20kHz so 44.1kHz is good enough for me. (You are limited to half the sample rate so at 44.1kHz the audio can't go over 22,050Hz and since filters are imperfect you can't get all the way too 22,050, but 20kHz is possible.)

"CD quality" is good enough for my tin ears... Most people can't hear the difference between a high resolution original and a copy down sampled to 16/44.1 in a proper blind, level-matched, ABX test. Often, you can't hear the difference between a high-resolution original and a good quality MP3, or you may have to listen very carefully to hear the difference. ...Again, that's assuming a controlled blind listening test.

that drive a class A amp in my stereo system,
Class-A isn't a "grade". ;) It's not necessarily better. It's simply a circuit classification. It was first (the oldest) amplifier circuit so it's "A". In the old days it was cheaper (for low power amplifiers) because the output stage only needs one tube or transistor. Modern solid state electronics is cheaper and adding one more transistor or MOSFET to an IC (for class A/B) adds nothing to the cost.

It's also the least energy-efficient which isn't a big deal with a headphone amp (or a preamp) that puts-out a few milliwatts but it can be an issue with a higher-power amp. Class-A wastes lots of power internally and it's basically wasting full-power even at low volumes or with silence.
 
Last edited:
OP
D

Deleted member 65547

Guest
Any measurements? Manufacturer's aren't always honest.

Any cheap DAC/soundcard can play a 32/384kHz file because the drivers will down-sample to match the hardware limitations. It's just like printing a high-resolution photo on a low-resolution printer. The manufacturer can semi-legitimately claim that it "supports" 32/384.

There are very few 32-bit DACs and most 24-bit DACs are only accurate to around 20-bits so you're never really going to get 32 useful bits (or need 32-bits).

I'm not sure what my upper limit is but I can't hear to 20kHz so 44.1kHz is good enough for me. (You are limited to half the sample rate so at 44.1kHz the audio can't go over 22,050Hz and since filters are imperfect you can't get all the way too 22,050, but 20kHz is possible.)

"CD quality" is good enough for my tin ears... Most people can't hear the difference between a high resolution original and a copy down sampled to 16/44.1 in a proper blind, level-matched, ABX test. Often, you can't hear the difference between a high-resolution original and a good quality MP3, or you may have to listen very carefully to hear the difference. ...Again, that's assuming a controlled blind listening test.


Class-A isn't a "grade". ;) It's not necessarily better. It's simply a circuit classification. It was first (the oldest) amplifier circuit so it's "A". In the old days it was cheaper (for low power amplifiers) because the output stage only needs one tube or transistor. Modern solid state electronics is cheaper and adding one more transistor or MOSFET to an IC (for class A/B) adds nothing to the cost.

It's also the least energy-efficient which isn't a big deal with a headphone amp (or a preamp) that puts-out a few milliwatts but it can be an issue with a higher-power amp. Class-A wastes lots of power internally and it's basically wasting full-power even at low volumes or with silence.
I haven't taken measurements; I consider my ears reliable enough. In any case, I prefer to listen through my eardrums rather than measurement tools, so it's crucial that they are satisfied. Additionally, I wouldn't stroll through the city wearing high-end headphones on my head, so between a honk and a sneeze, the listening difference with a portable DAC would be imperceptible anyway.

A DAC for mobile phones should sound good, nothing more, and these sound very good. You could invest the €20 needed to buy them, take some measurements, and then let us know...

As for Class A, you may be right, but being old-fashioned, I have a habit of biasing in Class A. And, having taken many measurements in this case, my spectrum analyzer says I did well.

What you say about the energy balance is certainly true for power amplifiers, where I am very satisfied with my Class D multi amplification, but for a headphone amplifier it is not so important, in the end it has a bias consumption of around 8 W.
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,406
Likes
18,370
Location
Netherlands
Chip will do about 75 to 95 SINAD depending on load and configuration. Nothing very special and I doubt it will be much better than the $10 Apple dongle.

A bit bigger but quite a bit better:

 
OP
D

Deleted member 65547

Guest
Chip will do about 75 to 95 SINAD depending on load and configuration. Nothing very special and I doubt it will be much better than the $10 Apple dongle.

A bit bigger but quite a bit better:

We would need to find the datasheet for the CS46L41.
 
OP
D

Deleted member 65547

Guest
From what can be gathered performance should be comparable to CS42L42. Dynamic range is a bit better though.
It seems very different to me, the series is the same but the cs46l41 is the higher version.
I would like to point out that the CS42L42 is a 24/192 while the CS46L41 is a 32/768... in short, very different
The characteristics declared by the manufacturer of the small DAC are:

  • [CS46L41 DAC ADAPTER]: The Cirrus Logic CS4641 is a HI-FI digital audio module with an integrated audio decoding chip. The excellent circuit layout and high-quality component selection make the entire module have excellent performance. Dynamic range up to 120dB, harmonic distortion less than 0.00125%

Obviously these are the characteristics of the chip implemented in an ideal circuit, which cannot be in a small dongle (which will obviously have jitter and power supply cleanliness problems), but as a starting point they are truly remarkable.

Anyway, the purpose of my original post wasn't so much about precise measurements as it was to highlight what modern nanoelectronics can achieve. Then, as mentioned earlier, for a DAC designed for mobile phones, whether it has 120dB of dynamic range and 0.00125% distortion or 100dB and 0.01%, it doesn't make much difference... I don't think anyone walks the dog with €5,000 HifiMan headphones on their head.
 
Last edited by a moderator:
OP
D

Deleted member 65547

Guest
My dear Jim, I have spent my whole life making measurements and perhaps this is why when I listen to music I prefer to concentrate on the pleasure it generates in me rather than on instrumental readings.
 
D

Deleted member 21219

Guest
My dear Jim, I have spent my whole life making measurements and perhaps this is why when I listen to music I prefer to concentrate on the pleasure it generates in me rather than on instrumental readings.

My dear @Uncino, that's putting it quite differently than your original post. Now I understand what you mean. Thank you. :)

Jim
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,406
Likes
18,370
Location
Netherlands
The characteristics declared by the manufacturer of the small DAC are:
Curious how the Cirrus product summary guide I linked has different specs listed.
 
OP
D

Deleted member 65547

Guest
My dear @Uncino, that's putting it quite differently than your original post. Now I understand what you mean. Thank you. :)

Jim
Why different? Since the first post I had said the same things, now I simply added that it's not that I lack measuring tools, it's just that I use them for more important things than knowing how many zeros after the decimal point the distortion of a DAC intended for listening to music has with the mobile phone.
I like walking while hearing Mother Earth Jazz (well), and I think this is more or less the need of everyone who connects headphones to the phone.
 
OP
D

Deleted member 65547

Guest
Curious how the Cirrus product summary guide I linked has different specs listed
Unfortunately I cannot find the complete datasheet, but I imagine that the manufacturer simply reported the data declared by Cirrus.
However it is not very strange, one is a series 42 and the other a 46
 
OP
D

Deleted member 65547

Guest
Curious how the Cirrus product summary guide I linked has different specs listed.
These are the complete features, I wonder why two different DACs... maybe the most convenient one is selected based on the characteristics of the stream and/or load?
Or maybe if the cell phone cannot use the CS46L41 the other one is activated?

[CS46L41 DAC ADAPTER]: The Cirrus Logic CS4641 is a HI-FI digital audio module with an integrated audio decoding chip. The excellent circuit layout and high-quality component selection make the entire module have excellent performance. Dynamic range up to 120dB, harmonic distortion less than 0.00125%
[PCM32Bit/384KHz]: Support 32bit 384Kh PCM decoding truly realizes Hi lossless music enjoyment. In addition, it also has good device compatibility for Huawei, Xiaomi, OPPO, VIVO, OnePlus. Google, Samsung, SONY, Compatible with more than 95% of mobile phone models, using UAC2.0 standard.
[DUAL DECODING]: High-Performance DAC digital chip CS46L41 + SA9302L dual Decoding chip 600 High Impedance independent ido power supply filter circuit capacitance
[AUDIO QUALITY]: Supports decoding and Playback of DSB64/128PCM mode, which can provide super quality signals from the beginning of the audio source effectively reducing the loss and distortion of the audio source for the use of DAC
 
OP
D

Deleted member 65547

Guest
I don't know about better, but the numbers are certainly bigger, yes :)
There may be doubts as to whether it is audible with speakers, especially if the reproduction chain is not of good characteristics, but in headphones the difference is always audible.
Additionally, DAC rolloff filters perform better by increasing the number of bits and sampling rate.
 
Top Bottom