Review: Apple vs Google USB-C Headphone Adapters

[slinkygn]

What I am more interested in is the built-in ADC. I plug it in using a TRRS connector and it works on Windows 10. Subjectively it sounds even slighter "richer" than SoundBlaster Play!4 and it would be great to confirm that objectively. @amirm considering how popular this dongle is, maybe you could throw some quick ADC tests like you did for SoundBlaster G6?

It would be nice to have broader tests for cheap, popular devices like these that can be potentially hacked into lots of different uses. Another interesting bit of data, I think, would be to get latency timing for the "cheap dongle" groups. Even if (depending on how the test was performed) the measure had to be relative, it'd be interesting to see those numbers -- the Google v2 dongle was marketed as having been intended to improve not sound quality, but latency, and to do so dramatically (which we now know seems to have come at the cost of sound quality); if one of the better performing dongles like the Apple or Samsung (or Meizu even, I suppose!) also had latency comparable or superior to the Google v2 dongle that'd be an interesting thing to know!

 

[Autechre]

Would you recommend a DAC if I already have a MacBook Pro 2020 and a dedicated amp? There is no 'noise' from the headphone jack as is, so would I benefit from an external DAC?

 

[meowchin]

Would you recommend a DAC if I already have a MacBook Pro 2020 and a dedicated amp? There is no 'noise' from the headphone jack as is, so would I benefit from an external DAC?

If you don't have any issues with audible noise, you won't really benefit from a dedicated DAC. A dedicated DAC might fare out better in measurements, but it's unlikely that you would be able to hear the difference.

 

[KeithPhantom]

Question: I have a Sanskrit 10th Mk.II as a dedicated DAC but it both heats up and uses battery from my iPhone even though it is in data-only mode (the DAC itself is powered by an external power supply). I am using the Lightning-to-3.5mm adapter as a DAC, but it only outputs 1 Vrms. My amplifier for now is the A50s, which I use in high gain with the balanced connection since there’s almost no penalty for using it like this versus low gain and I can power my HD 600 to my normal listening levels at 25% of the volume knob. I already know the Apple dongle measures well and everything else’s fine, but can I keep using this dongle as a DAC or it is advisable to buy another one for any other thing that I may be missing?

 

[3125b]

Here is my measurement of the European Apple dongle (A2155):

With a -1dBFS 1kHz signal the output voltage is 469mVrms (461mV measured with TrueRMS multimeter), SINAD is a diappointing 86dB.
Measurment taken via RME ADI-2 Pro FS.

 

[MOCKBA]

Mind you, built-in DAC in LG G7 ThinQ is much better still so Android fans don't need to kill themselves.

I have LG G8 quad DAC, and it is even not in you list. I wish to switch to iPhone if they add MicroSD card support. I think it is a good time, since Samsung dropped it.

 

[cenix]

Hello guys,

I have been following this for a bit and bought the dongle eventually. I have the EU version (A2155) and the volume seems pretty low on Windows 10. I read in here that the US version has twice the output compared to the EU version, is this true across all Windows apps (including Foobar2000)?

Or, do the EU- and US versions have the same output, but they are somehow blocked by some application?

I just want to know if it's software or hardware related, in terms of the difference in output.

Thank you.

Edit: Does anyone know where to buy the US version from Europe? I looked on Ebay for a bit, but can't find one without a huge shipping price.

 

[SDX-LV]

Here is my measurement of the European Apple dongle (A2155):
With a -1dBFS 1kHz signal the output voltage is 469mVrms (461mV measured with TrueRMS multimeter), SINAD is a diappointing 86dB.
Measurment taken via RME ADI-2 Pro FS.

Interesting result. I certainly did not measure my equipment, but I also did notice any issues with my A2155 (produced 12/2020) as with my older ODAC+O2 combo (same same).
Are you sure the measurement is comparable? This means that Apple dongle got worse, or what could be an explanation?

 

[3125b]

@SDX-LV
I actually have to correct myself there. That measurement was taken using the same PC as a source that was connected to the RME via USB. Using the dongle on my laptop made a big difference. The result seems to stem either from a electrical potential issue to wich the Apple dongle seems to be particularly susceptible or the quality of the USB signal (or supply voltage since the dongle is made to run on battery powered devices). I have noticed barely any difference in the FiiO E10K for example.
Here is the new measurement, wich is about a dB better than even Amirs measurement of the US version (at twice the voltage no less):


Again, I'm not sure what the cause of the problem is, but maybe it is the same (or even worse) in your situation. Under certain conditions, the EU dongle measures very well given its low output voltage.

 

[SDX-LV]

@SDX-LV
The result seems to stem either from a electrical potential issue to wich the Apple dongle seems to be particularly susceptible or the quality of the USB signal (or supply voltage since the dongle is made to run on battery powered devices).

This could make sense. As you say these dongles are normally powered by clean battery power from iPad. And there is not much room for power filtering or anything in the minimalistic dongle format. So this is reproducible then, Apple dongle is <90 SINAD on one PC and ~100 on another?

This adds a major uncertainty in what performance we get in PC application when plugged into the wall.

 

[3125b]

I have done a few other tests, that particular problem seems to be mostly due to potential issues.
Here is a measurement at full output (with a -1dBFS signal as always, that seems to work fine with a devices tested so far) via the laptop running on battery (as was the case in the measurement above) with 16bit selected in Windows:

92dB SINAD is not far from the theoretical maximum of 96dB with 16bits. Not too bad.

And here is a 50mV measurement with the laptop power supply plugged in:

First of all, 87dB SINAD @50mV places it in the 1/3 of all devices tested by Amir, so that is not bad at all, very respectable for a 10€ dongle. But you can see the spikes at 50 and 100Hz. So the dongle does seem to have less than ideal filtering. However the noise floor is very low here, those spikes are not concerning at all.

I will probably repeat the test (and maybe others) with the dongle (and many other devices) plugged into the tower PC when I set the RME up with my laptop in the future. I have to figure a few things out, testing via ADC in REW is nowhere near as easy and versatile as using an analyzer, but I will try my best to get ahold of a good selection of budget devices to test because that's what people buy most and where you are most likely to find bad apples (no pun intended). I might be setting up a website in the future as a college project, we'll see.

 

[Ismapics]

This is my recommended before and After

I use simple headphones and the news that this dongle gives great performance is encouraging. So in order to get more longevity I decided to heat shrink wrap it to help with strain. check out the pics. Actually happy with the dongle!

 

[sadburai]

hi guys,
is it possible to use multiple apple usb c adapters on the same windows pc?

I'd like to connect both my stereo speakers and my headphones to an adapter each and then select the output device from the windows sound menu

 

[Veri]

hi guys,
is it possible to use multiple apple usb c adapters on the same windows pc?

I'd like to connect both my stereo speakers and my headphones to an adapter each and then select the output device from the windows sound menu

Sure, don't see why that won't work. You'll just have to change active device is all. Just note you need Windows 10. 7/8 don't have the general purpose driver to interface to it.

 

[CuteStudio]

I was able to wire it to a Raspberry Pi Zero W through a USB 2.0/Type C adapter:

I was also able to get it working (ona Pi3b) with a one piece USB-A to USB-C adapter.
I struggled to get the Pi to see it as an audio device until I plugged something into the 3.5mm socket: then it wakes up, lesson learned

I (basically) use aplay to drive it in 24bits.
It's an amazing DAC. Also the smallest I've even (not quite) seen !

 

[sadburai]

Sure, don't see why that won't work. You'll just have to change active device is all. Just note you need Windows 10. 7/8 don't have the general purpose driver to interface to it.

Works well! So great to finally control headphones and speakers individually, compared to my FiiO E10. And the Windows 10 sound drivers even allow to play sound from different apps on different audio devices, e.g. Spotify to speakers, chrome/youtube to headphones.

I have a shortcut on my desktop to change the app volumes for the different devices. I think it's quite handy:
1.) Create a new shortcut on desktop with right-click > New > Shortcut
2.) change URL of shortcut to: ms-settings:apps-volume
Unfortunately, you cannot pin it to the taskbar or start menu.

 

[HITMAnsOFT]

I was also able to get it working (ona Pi3b) with a one piece USB-A to USB-C adapter.
I struggled to get the Pi to see it as an audio device until I plugged something into the 3.5mm socket: then it wakes up, lesson learned

I (basically) use aplay to drive it in 24bits.
It's an amazing DAC. Also the smallest I've even (not quite) seen !

I read beforehand that it has to have something plugged in to be detected and that's a minor inconvenience, but it seems to be common with highly integrated dongles. For some reason I could not get the USB on RPi Zero work with Volumio 2.x stable version, so I'm forced to use 3.x beta.

I'm running Volumio on this portable player project, mostly as an AirPlay receiver, but sometimes as a DLNA client. The Now Playing screen on the Volumio shows 24bit, so that should be good.

This is definitely an improvement over my previous no-name PCM5102 PCB that simply is not designed to directly drive headphones.

 

[CuteStudio]

I think the 'don't be an audio adapter unless something is plugged in' feature is actually to save battery power when on a laptop etc.

I used mine into the gate of a triode, so there's not such a load on it
I used to use a couple of large boxes for rate conversion and DAC, but I sold those to make space, and moved to a PCM2704 (the cheapest I could find, but surprisingly good), which has a nice direct sound:

To the Apple USB-C headphone dongle, which is only 50% more (£7.50) but definitely sounds more realistic. On several tracks I can hear a more real, lifelike image, so it's definitely a keeper.

I have no idea how Apple fit all the stuff needed for a 24 bit version of the above, with ADC attached, into basically a USB plug, and make it sound so good, but it is very good.

Ironically it doesn't actually work on my Apple airbook, but in the Pi3 it's very sweet. the Pi is on the regular PSU from the Pi shop and works just fine, so it's a very compact audio slave.

 

[edechamps]

@SDX-LV
I actually have to correct myself there. That measurement was taken using the same PC as a source that was connected to the RME via USB. Using the dongle on my laptop made a big difference. The result seems to stem either from a electrical potential issue to wich the Apple dongle seems to be particularly susceptible or the quality of the USB signal (or supply voltage since the dongle is made to run on battery powered devices).

Unless you took special measures to isolate the test setup, it's extremely likely you just had a ground loop. It's a very typical problem in "homemade" measuring setups. You can avoid it by using an USB isolator for example.

 

[DSJR]

I can't use the main rig much, so music via this ancient Dell laptop is the thing currently. being a pauper cheapskate, I bought the EU version of the Apple USB C dongle (from UK suppliers AO) for £7 and then a C-A adaptor from eBay UK for just over two quid.

I use lower mid level Sennheiser 'phones at 64 ohms or thereabouts and there's 'just' enough output level for gently spirited listening and there seems enough to feed my preamp quite well if windows volume is maxed. Via headphones, I have to say the clarity of sound and bass quality does seem better than the headphone output of the Dell, which is a simpler core 2 model from Jurassic times, albeit with upgraded processor and maxed memory as befits the mobo capacity.

Got to say that if the US version does have higher output, I'd be keen to try one as long as shipping and duties don't hobble the price.