Is wishbone bus I2C or SPI?


#1

Hi all,

I’m doing some testing and research about beamforming on microphone array. And I’m wondering the wishbone bus is I2C or SPI? I didn’t find any documentation about that.

What’s more, when I did same testing experiments to MATRIX and try to find the phase delay between each microphone, the audio result is not the same. That’s why I’m thinking to kill all the unnessary process ran in the background. Do you have any idea about which process I can kill? (In other word, is there any other process running on MATRIX?)

Thanks.


#2

Hi @GoodLife,

The wishbone bus is connected to the Raspberry Pi using SPI.

when I did same testing experiments to MATRIX and try to find the phase delay between each microphone, the audio result is not the same.

Regarding this questions. Can you elaborate about the experiments, and the audio result ?

Also wanted to know how are you getting the audio from the mics, from alsa or using HAL demos ?

Yoel


#3

I use HAL demo (micarray_recorder), and play a single tone from phone above the center of MATRIX. I plot the result from each mic into matlab, try to capture single period of the single tone among all mics, and use the position of each peak value to determine the relative phase difference among all 8 microphones. But the thing is, different day’s recording will give different result. So I’m wondering maybe there’s some other process is using wishbone bus. @yoelrc88


#4

If you are using HAL example you have to make sure MALOS is not running at the sane time. The both HAL and MALOS talk to the MATRIX Creator using SPI, and when the run at the same time the audio signal gets some errors. So , try killing MALOS services using : sudo pkill malos

Yoel


#5

yep I already kill malos, but as I memtioned before, the result is not consistent. Is there any initialization done by wishbone bus? Or is there any other process is using SPI?
@yoelrc88


#6

No, normally after killing malos the mic examples from HAL work fine. Can you record and try to “see” the audio signal. Maybe you can share the audio files so we can review them.

Yoel


#7

I’m trying to do the experiment again, if this scenario still happens I will report here