Audio problems with a 16-Channel board based on the XUF216-512-TQ128

[jwo]

Hello,

I hope this is the correct place to post this.
I'm using a board with a XUF216-512-TQ128 processor, which is based on the xCORE Array Microphone Board but with 16 MEMS microphones (IM69D120). The firmware is based on the app_usb_aud_mic_array from the XMOS USB Audio 2.0 Software package.
The problem is, that I get some random sizzling, whining, high pitched, modulating sounds. They occur randomly on all channels, sometimes at the same time on all channels. Attached is a spectrogram of one channel where it is occuring. It happens with all samplerates and also with the original firmware (of course then there are only 7 mics available).
I was wondering, if anyone has experienced something similar or has an idea what this could be.

Thanks in advance for any help,
best regards

[mon2]

Hi. Has this board ever worked correctly? Is the board an official XMOS kit or a custom design? If XMOS kit, which exact p/n?

You can try to review the signals coming out of the microphones and while tapping the microphone - do you see related outputs? On this forum there are other developers noting that the issue could be with the soldering of the microphone(s).

https://www.xcore.com/viewtopic.php?f=8 ... ise#p33855

* different CPU but do review
https://www.xcore.com/viewtopic.php?f=4 ... ise#p31840

[akp]

Sounds like this is a custom board. Have you tried it on the XMOS board you referenced?

[jwo]

Thanks for your answers. It is a custom design, although not by me. I will definitely try tapping the microphones while recording and report the results I got.
It could very well be a soldering problem. Every mic is soldered onto a single small PCB which is connected to the main board via ribbon cables. I tried different cable lengths with no difference, so I would source out the cables as the cause. The original XMOS mic array board works as expected. The microphones on that board are the IM69D130 compared to the IM69D120 I'm using, but I don't think that should make a difference, right?
Thanks again, I'll report once I ran further tests. If anyone has some further ideas please let me know.

[jwo]

I just tested holding my finger on the back of the mems when the sound occurs and that does definitely influence the sound.
It seems like while my finger is on the back of the microphone the sound is modulating much faster but is also a lot quieter, from what I could observe during first tests.
What is striking me as a bit odd is that the same phenomenon occurs with all microphones, which would mean that they all have the same fault.