Let me return to the EMI issue with this amplifier. The operation of this amp has made Internet 2.4GHz modem/Wi-Fi totally unusable, as I already wrote in this thread. I was prepared to throw the amp to household waste recycling site. But I gave it one more chance finally.
Infeon application note AN_2105_PL88_2105_151947 says this:
"The radiated emission test is passed with a margin better than 5 dB over the whole frequency range of 30 MHz to 1 GHz with speaker cables up to 40 cm (see section 3). It is also possible to pass radiated emission with 60 cm cables, though with less margin than
with 40 cm cables."
For cables longer than 60 cm they recommend this:
So I have prepared an additional filter to check the situation
Nothing very nice, but enough to try. I have used my my circular wire probe (placed near the speaker wires) that measures magnetic flux intensity dB/dt and compared results with and without the LC filter.
1. Without the LC filter
Peak-to-Peak induced voltage is 49.5 mV.
2. With the LC filter (33uH, 470nF)
The Peak-to-Peak induced voltage has dropped to 6.34 mV and there is no issue with Wi-Fi anymore. The circular probe is shown just for illustration, dB/dt was measured near the speaker cable.
Conclusion
I do not believe that the SMSL-A100 has passed any EMI radiation test, without additional filters and with cables. To me, any info about this amp passing the test is either misleading or fake.
It is possible to make it working without unwanted radiation, but then it is not a "filterless" amplifier.