- Joined
- Oct 10, 2020
- Messages
- 806
- Likes
- 2,638
FiiO Taishan D03K is a pretty cheap DAC that was often recommended online to connect TVs (that now mostly provide only digital audio output) with analogue audio gear. I have one so I thought it might be useful to post my measurements. Cost was 21,99 GBP a few years ago.
Here's a few pictures of the device:
The measurements you'll see next were done with the RME Babyface soundcard, first generation - measured here. The DAC was connected to optical SPDIF output of RME Babyface, clock source was RME internal clock and the sample rate was set to 48kHz for most measurements. Also, in most cases the Hi-Z unbalanced input of RME Babyface was used to measure the DAC performance. Exceptions will be noted in the text.
Let's start with the basics, RMAA measurement summary:
We can see that left channel performs better that the right, but both are far from SOTA. Still, should be more than OK for the intended uncritical use.
Note that I had to measure each channel individually as I have only one high-impedance (Hi-Z) unbalanced input on my soundcard. To measure channel crosstalk I had to use the stereo unbalanced inputs (which are relatively low-impedance in my soundcard). Using low-impedance inputs with this DAC caused excessive loading of its output and as a consequence would increase distortion - which is why I didn't use them for other tests.
Here's the frequency response:
Not a lot of channel level difference, but there is a ~0,9dB rise cantered around 20kHz. Note that the above response is calibrated/normalized to my soundcard's loopback response (which should be dominated by ADC response, based on my analysis here). This calibration is what causes the sharp peak in the FR extreme at ~23kHz (the part of the response with a very steep slope), and not the DAC under test, so it can be disregarded.
If we change the sample rate to 96kHz we see a similar rise past 20kHz (again, disregard the sharp peak above ~40kHz):
Next let's see the spectrum of a 1 kHz sine at -1 dBFS peak output:
Reasonably clean noise spectrum, but SINAD is only ~91dB. Maximum full scale sine output is ~1,53 Vrms.
Let's look now at the THD vs level sweep at 1kHz:
We see that distortion starts to emerge from the noise floor between -30 and -25dBFS, and highest SINAD is reached at -1 dBFS peak (-4 dBFS RMS).
Let's look at the THD vs frequency at -1dBFS (1,37Vrms) output:
Here we can again see the noise level difference between channels.
And THD comparison at 0,5 Vrms vs 1,37 Vrms output levels:
We can see that the 2nd and 3rd harmonic dominate THD, and we see also a significant rise in THD towards the lowest frequencies. Other than that nothing very exciting.
Multitone at -1dBFS output:
Definitely not great
(~12 bits distortion-free)
Jitter:
Some artefacts are visible but all are below -110dB.
In conclusion I'll just say that, while D03K doesn't measure very well, it is tiny and IMHO does the job reasonable well for uncritical purposes. E.g. I used it with my TV for a while, and recently only use it as a tool when I'm having issues with ground loops in certain setups (which can be broken with the optical connection to the DAC). I wouldn't use it as a signal generator to measure SOTA ADCs, though!
Here's a few pictures of the device:
The measurements you'll see next were done with the RME Babyface soundcard, first generation - measured here. The DAC was connected to optical SPDIF output of RME Babyface, clock source was RME internal clock and the sample rate was set to 48kHz for most measurements. Also, in most cases the Hi-Z unbalanced input of RME Babyface was used to measure the DAC performance. Exceptions will be noted in the text.
Let's start with the basics, RMAA measurement summary:
We can see that left channel performs better that the right, but both are far from SOTA. Still, should be more than OK for the intended uncritical use.
Note that I had to measure each channel individually as I have only one high-impedance (Hi-Z) unbalanced input on my soundcard. To measure channel crosstalk I had to use the stereo unbalanced inputs (which are relatively low-impedance in my soundcard). Using low-impedance inputs with this DAC caused excessive loading of its output and as a consequence would increase distortion - which is why I didn't use them for other tests.
Here's the frequency response:
Not a lot of channel level difference, but there is a ~0,9dB rise cantered around 20kHz. Note that the above response is calibrated/normalized to my soundcard's loopback response (which should be dominated by ADC response, based on my analysis here). This calibration is what causes the sharp peak in the FR extreme at ~23kHz (the part of the response with a very steep slope), and not the DAC under test, so it can be disregarded.
If we change the sample rate to 96kHz we see a similar rise past 20kHz (again, disregard the sharp peak above ~40kHz):
Next let's see the spectrum of a 1 kHz sine at -1 dBFS peak output:
Reasonably clean noise spectrum, but SINAD is only ~91dB. Maximum full scale sine output is ~1,53 Vrms.
Let's look now at the THD vs level sweep at 1kHz:
We see that distortion starts to emerge from the noise floor between -30 and -25dBFS, and highest SINAD is reached at -1 dBFS peak (-4 dBFS RMS).
Let's look at the THD vs frequency at -1dBFS (1,37Vrms) output:
Here we can again see the noise level difference between channels.
And THD comparison at 0,5 Vrms vs 1,37 Vrms output levels:
We can see that the 2nd and 3rd harmonic dominate THD, and we see also a significant rise in THD towards the lowest frequencies. Other than that nothing very exciting.
Multitone at -1dBFS output:
Definitely not great
(~12 bits distortion-free)Jitter:
Some artefacts are visible but all are below -110dB.
In conclusion I'll just say that, while D03K doesn't measure very well, it is tiny and IMHO does the job reasonable well for uncritical purposes. E.g. I used it with my TV for a while, and recently only use it as a tool when I'm having issues with ground loops in certain setups (which can be broken with the optical connection to the DAC). I wouldn't use it as a signal generator to measure SOTA ADCs, though!
Last edited:
