Dell CN-03GVVD USB-C Headphone Adapter Review

[TheWalkman]

For the price of a couple of packs of gum, you get to enjoy music out of your phone.


I’ll bank my chewing gum allowance and splurge for the Apple dongle.

 

[S=klogW]

You might want to redo the dashboard anyway. It shows .49V output, but the impedance sweep and the power output for both 300 Ohm and 32 Ohm all clearly show that it can do just shy of 1V. So something isn't matching up and I suspect your initial dashboard is wrong in more than just the one-channel issue.

debunker, you beat me to it. Power=Vsquared/R is one of my favourite equations. I too vote 0.98Vrms for the Dellongle

 

[KxDx]

I understand selling loudspeakers in pairs but this ... How do you use a pair of these?

I have 3 headphone dongles... I keep a pair of headphones at home with my exercise bike, and a pair in each car so when I go walking I have headphones available.

My short term memory is bad enough that I don't trust myself to remember my headphones when I leave home, so I keep a pair in each car!

 

[beefkabob]

I rarely use headphones with wires, but when I do, I can never find the dongle. It's better to just leave one of these attached to each pair of headphones.

 

[pseudoid]

Thank you for the review, @amirm,
How odd:

The Dell Adapter - USB-C to 3.5mm Headphone Jack enables you to connect to devices that use a standard 3.5mm audio plug. With this adapter, you can easily connect your laptop or tablet with a USB Type-C port to any devices (headsets or speakers) that use a standard 3.5mm audio plug. Take it with you wherever you go. Just place it in your bag, and head off to the next meeting or conference room. Get started right away with the build-in driver for easy set-up.

Nowhere (neither herein or at amazon), do I see that this is a 'stereo' RCA 3.5mm audio output.
Although it should be a given, amazon offering a pair makes it suspicious!

 

[Rja4000]

but the impedance sweep and the power output for both 300 Ohm and 32 Ohm all clearly show that it can do just shy of 1V.

Which explains why power could be subjectively "OK".
With 0,5V, I think, personally, it's not enough.

 

[Rja4000]

I then decided to run another test with AP driving it and it produced both channels.

That's interesting.
And it probably explains why you were able to provide more extensive testing, doesn't it ?
I guess this doesn't work with the Apple dongle ?

 

[Rja4000]

all dongles regardless of cost:

I still wonder why the Tanjchim Space doesn't show here ?

 

[amirm]

That's interesting.
And it probably explains why you were able to provide more extensive testing, doesn't it ?
I guess this doesn't work with the Apple dongle ?

With many of these devices, the ASIO emulator I use screws up the path so I can't use the AP to drive them. Here, it worked well. Maybe something has been fixed in Windows to allow it.

 

[amirm]

I still wonder why the Tanjchim Space doesn't show here ?

I had forgotten that I had that separate graph for dongles at that time. I will add it if I remember....

 

[lewdish]

I miss these goofy dongle dac reviews, L7 used to do tons of dongle reviews but he hasnt posted in years now.

 

[a2Thompson]

I rarely use headphones with wires, but when I do, I can never find the dongle. It's better to just leave one of these attached to each pair of headphones.

All I have our wired headphones for my phone and so I also have an equal number of dongles that stay attached to them.

 

[beefkabob]

If this is the same adapter: https://qzxx.com/79408.html


…it would be the fairly common CX31993 chip.
SA1023z, HHW59, 750-BBDJ appear to be other Dell pn of the same (?) adapter.

CX31993’s brother, the balanced CX31988, was reviewed as part of the Yongse Y01: slightly better performances and (coincidentally?) exact same 3.2 / 29 mW measured power.

It does have that label on the metal, so you're correct, I assume.

 

[CedarX]

It does have that label on the metal, so you're correct, I assume.

I ordered a set on Amazon (ASIN B0CCSWP866), and this what I received (in a generic envelope itself packed in the Amazon envelope):

Although they have different marking, I didn’t see any difference when connected to my laptop.
Next step: cut one open to confirm the DAC/Amp chip…

 

[beefkabob]

This is what amirm tested.

 

[markanini]

I wonder how the output voltage was limited to 0.5V, if it's CX31993-based it should be capable of 1V.

 

[AnalogSteph]

But I am not sure why it cuts off at 18 kHz:

At 48 kHz, 18 kHz becomes 19.6 kHz and 21.9 kHz becomes 23.8 kHz. Perhaps someone accidentally designed a filter with cutoff frequencies for 44.1 at 48 kHz instead? In any case they were arguably making life needless hard for themselves by going for cutoff even slightly below fs/2 and better than -90 dB.

Periodic filter ripple seems to be about ±0.025 dB, neither great nor terrible. Assuming an FIR filter, I would consider this marginal. Hi-Fi grade converters tend to be sub ±0.005 dB or so, and while that is fine, standards in this regard have actually gone down in the last 15 years (it used to be not uncommon to see values of ±0.0000xx dB, and even Realtek eventually caved in and released chips with ±0.0005 dB on the DAC side).

The combination of both doesn't exactly make for the most balanced filter performance. (It would do quite well in a consumer ADC at 48 kHz and up.) I'm kind of not surprised that this chip (assuming it's the CX31993) only ever seems to grace USB dongles.

 

[Jimster480]

These little adapters are quite good these days. I remember when performance like this was found in actual desktop products with discrete power supplies and such

 

[CedarX]

So, I finally took the time to cut one open:

It is a Synaptics (ex-Conexant) CX31993 chip…

 

[CedarX]

I wonder how the output voltage was limited to 0.5V, if it's CX31993-based it should be capable of 1V.

I have a few other generic CX31993 adapters that I need to cut open as well. These chips have an on-board ARM processor (Cortex M0?). The 0.5V could be a firmware setting…

Another explanation (speculation…) could be a charge pump limitation depending, for example, on the inductor being used.
I have noticed that the Moondrop FreeDSP, which is most likely built around a CX31988 has a “large” (relatively speaking…) discrete component on the rear side, assumed to be the charge pump inductor.
The 5V USB “+” rail is not going to be the limitation for 0.5 or 1V…