Hi all,
I've just completed a revision to a board using the AIC3204 with a CC1120 radio and C1190 LNA/PA and I have an issue with what appears to be noise coupling into the microphone input and being converted by the codec.
The codec digital stream which contains the sampled microphone input audio contains "spike-like" noise mixed-in. This spike-like noise seems to coincide with the timing cycles of the RF transmissions of the CC1120/CC1190 RF circuit on the board. There are RF transmissions every 100ms and the spikes in the digitized audio codec signal is cyclic with 200ms cycles.
The other thing is that touching the antenna, and/or moving the relative position of the board closer to the mic-input cable (which is basically an off-the-shelf mono headset), increases/varies the digitized noise-spikes being sampled. This makes me think its definitely some kind of coupling of RF impulses into the codec. When I remove the head-set and touch the ground "ring" of the codec input jack, the 100ms cyclic noise-spikes increase in amplitude and are very evident at the audio output as a "chopping" type of sound...
My question is, how to resolve this apparent coupling of an UHF RF transmission into the codec ? I have for instance placed a 100pF NPO shunt cap to ground on the mic input, but this does not change much. Perhaps it requires a larger, say 100nF shunt, or effectively a low-pass filter on the codec input to filter these spikes out ? But I'd prefer avoiding the RF coupling at the source, rather thant to cut down the bandwidth of the codec input (currently sampling rate is set at 8KHz).
How can 915MHz RF signals couple into the codec digitized audio ? Is the path via the codec GND ground, or perhaps through the codec power supply 3v3 ? Would isolating the codec power pins and its analogue ground with ferrite-beads, away from the rest of the circuit supply and digital/RF ground, help in removing this annoying EMI/RFI ? Would a nickel/copper shield over the RF circuit reduce or eliminate the RF coupling into the codec for instance ? What are the most commonly pursued/recommended solutions to resolving EMI/RFI audio noise/RF coupling problems ?? Unfortunately the AIC3204 reference/evaluation design does not provide much insight into this type of problem ...
Thanks in advance for any suggestions/ experience with similar issue.
Regards Mikl
