- Joined
- Oct 10, 2020
- Messages
- 859
- Likes
- 2,766
Let me share a few basic measurements of the Arylic A50+ streaming amplifier. The amp is based on the TPA3116 chip and I bought mine together with the ACPWorkbench Tuning Tool app for $152.32 total (shipped to EU) during the holiday sale. A few pictures of the amp:
The included laptop-style PSU looks quite bulky compared to the tiny amp:
The included remote is very easy to use and the functions are clearly labelled. Many of the same functions (and additional functions) can also be accessed with the 4STREAM mobile app.
One thing I disliked from UX perspective is the fact that the amp itself doesn't have a display. That would be very useful in a device as feature rich as this one; and especially it would be useful to be able to see the volume setting at all times. Interestingly, the similar Arylic S50 Pro+ streaming preamp has a display (but no loudspeaker amp section).
The tools used for measurements:
Let's dive in!
1kHz SINAD
As we can see, unfortunately the performance is quite bad, especially in the right channel. SINAD is dominated by distortion (mainly 3rd harmonic), but noise is also quite high. More on noise later.
I tried measuring a few other inputs, but none of them gave better results - here's left channel performance when driven from RCA, USB and optical inputs:
Multitone response
Here's the multitone response (5W into 4Ohm, one channel driven):
This is not a good result - ideally we'd want better than -98dB / 16bits of resolution.
Frequency response and load (in)dependence
We can see that the amplifier is load-sensitive (indicative of high output impedance), and we also see that the channel matching is not ideal (and also load sensitive). The device digitizes even the analogue input and processes internally at 44,1kHz, so FR drops like a brick just before 20kHz. It seems it might the possible to change the sampling rate in the ACPWorkbench app, but even so it caps-out at 48kHz. Otherwise the FR is reasonably flat with these resistive loads. Using a complex load like a loudspeaker would introduce variations in the FR (see an explanation here).
Based on the measured voltage with 4 Ohm and 8 Ohm resistive loads I calculated and plotted the output impedance:
As we can see, left channel output impedance at 1kHz measures 0,68 Ohm and right channel 1,1 Ohm. This corresponds to left channel 8Ohm damping factor of ~12 and right channel ~7. Unfortunately that is quite a bad result.
Power into 4Ohms (20kHz BW, one channel driven)
Power into 8Ohms (20kHz BW, one channel driven)
Good news is that both corresponds to the manufacturer-specified power ratings.
DSP / PEQ capabilities
While the amplifier allows the user to set basic tone EQ (bass / treble) from the remote, the optional add-on ACPWorkbench Tuning Tool Windows app (price: 20$) allows the user to individually set up to 8 bands of PEQ for the left, right, sub channels, and an additional one for the overall stereo output - as well as to add various other processing (sub delay, source pre-gain, etc..).
Here's a snip of PEQ settings from the tool and the resulting FR measurements:
I thought it was very cool that the sub output has a dedicated DSP channel - that gives quite a bit of additional flexibility for subwoofer integration!
BTW, when you disable all DSP for the sub output (marked as 'mono channel' in the ACP app), there is still a 240Hz -12dB/oct LPF applied (and a strange, slow LF droop that reaches -7,6dB by 20Hz):
How does it work in practice?
I found the A50+ was very nice to use. The remote was very intuitive and the 4STREAM app worked well in the short time I tested it. It also works well with the WiiM mobile app. The feature set is IMHO really amazing, and especially for the low asking price.
However, when I connected the amp to my Revel M16 speakers, at 2,3m listening distance I could right away notice slight hissing/humming. When I came closer to the loudspeaker the noise was quite apparent, and its spectrum/character was changing depending on the status of the amp.
E.g. immediately after playback it is typical white noise spectrum hiss (I will call this the "Ready" status). After a few seconds it modulates to a hum (I will call this "Standby 1" status) and if you wait a bit longer the hum becomes much less audible (I will call this "Standby 2" status).
I noticed this behaviour while doing measurements as well - here's an example of how this looks like:
Apart from this, I'd rather not comment on the sound, because I didn't use any controls in my listening. It played IMHO loud enough in an apartment building context at 2,3m listening distance with 86dB (2.83V @ 1M) sensitivity bookshelf loudspeakers.
Conclusion
Really hard to rate this device - on one hand the feature set is quite unique for the very low price. On the other hand, measured performance is poor in most tests.
If the amp had a basic display and measured performance similar to at least Infineon/Merus-based amps it would be a winner. As is I can only recommend it for use-cases where audio performance is of secondary importance.
Thanks for reading!
The included laptop-style PSU looks quite bulky compared to the tiny amp:
The included remote is very easy to use and the functions are clearly labelled. Many of the same functions (and additional functions) can also be accessed with the 4STREAM mobile app.
One thing I disliked from UX perspective is the fact that the amp itself doesn't have a display. That would be very useful in a device as feature rich as this one; and especially it would be useful to be able to see the volume setting at all times. Interestingly, the similar Arylic S50 Pro+ streaming preamp has a display (but no loudspeaker amp section).
The tools used for measurements:
- RME Babyface Silver Edition as signal generator (USB input, Optical SPDIF output), 44,1kHz sample rate and 24bit depth
- E1DA Cosmos ADC (2,5mm TRRS Aux input), 44,1kHz sample rate and 32bit depth
- Asio4All drivers (44,1kHz sample rate)
- REW software for measurements
- Google Sheets for impedance calculations
- DIY resistive dummy load
- Revel M16 loudspeaker for listening
Let's dive in!
1kHz SINAD
As we can see, unfortunately the performance is quite bad, especially in the right channel. SINAD is dominated by distortion (mainly 3rd harmonic), but noise is also quite high. More on noise later.
I tried measuring a few other inputs, but none of them gave better results - here's left channel performance when driven from RCA, USB and optical inputs:
Multitone response
Here's the multitone response (5W into 4Ohm, one channel driven):
This is not a good result - ideally we'd want better than -98dB / 16bits of resolution.
Frequency response and load (in)dependence
We can see that the amplifier is load-sensitive (indicative of high output impedance), and we also see that the channel matching is not ideal (and also load sensitive). The device digitizes even the analogue input and processes internally at 44,1kHz, so FR drops like a brick just before 20kHz. It seems it might the possible to change the sampling rate in the ACPWorkbench app, but even so it caps-out at 48kHz. Otherwise the FR is reasonably flat with these resistive loads. Using a complex load like a loudspeaker would introduce variations in the FR (see an explanation here).
Based on the measured voltage with 4 Ohm and 8 Ohm resistive loads I calculated and plotted the output impedance:
As we can see, left channel output impedance at 1kHz measures 0,68 Ohm and right channel 1,1 Ohm. This corresponds to left channel 8Ohm damping factor of ~12 and right channel ~7. Unfortunately that is quite a bad result.
Power into 4Ohms (20kHz BW, one channel driven)
Power into 8Ohms (20kHz BW, one channel driven)
Good news is that both corresponds to the manufacturer-specified power ratings.
DSP / PEQ capabilities
While the amplifier allows the user to set basic tone EQ (bass / treble) from the remote, the optional add-on ACPWorkbench Tuning Tool Windows app (price: 20$) allows the user to individually set up to 8 bands of PEQ for the left, right, sub channels, and an additional one for the overall stereo output - as well as to add various other processing (sub delay, source pre-gain, etc..).
Here's a snip of PEQ settings from the tool and the resulting FR measurements:
I thought it was very cool that the sub output has a dedicated DSP channel - that gives quite a bit of additional flexibility for subwoofer integration!
BTW, when you disable all DSP for the sub output (marked as 'mono channel' in the ACP app), there is still a 240Hz -12dB/oct LPF applied (and a strange, slow LF droop that reaches -7,6dB by 20Hz):
How does it work in practice?
I found the A50+ was very nice to use. The remote was very intuitive and the 4STREAM app worked well in the short time I tested it. It also works well with the WiiM mobile app. The feature set is IMHO really amazing, and especially for the low asking price.
However, when I connected the amp to my Revel M16 speakers, at 2,3m listening distance I could right away notice slight hissing/humming. When I came closer to the loudspeaker the noise was quite apparent, and its spectrum/character was changing depending on the status of the amp.
E.g. immediately after playback it is typical white noise spectrum hiss (I will call this the "Ready" status). After a few seconds it modulates to a hum (I will call this "Standby 1" status) and if you wait a bit longer the hum becomes much less audible (I will call this "Standby 2" status).
I noticed this behaviour while doing measurements as well - here's an example of how this looks like:
Apart from this, I'd rather not comment on the sound, because I didn't use any controls in my listening. It played IMHO loud enough in an apartment building context at 2,3m listening distance with 86dB (2.83V @ 1M) sensitivity bookshelf loudspeakers.
Conclusion
Really hard to rate this device - on one hand the feature set is quite unique for the very low price. On the other hand, measured performance is poor in most tests.
If the amp had a basic display and measured performance similar to at least Infineon/Merus-based amps it would be a winner. As is I can only recommend it for use-cases where audio performance is of secondary importance.
Thanks for reading!
Last edited: