Which verified that the respeaker mic array is compatible with jetson nano. My problem is that I don’t know how to properly connect them. I can’t connect it like mentioned on the guide( reSpeaker 4-Mic Array for Raspberry Pi | Seeed Studio Wiki) because it doesn’t fit, the ethernet socket is too close to the GPIO header. I am a beginner at developing on the jetson nano and embedded devices in general. I would appreciate any advice
You’d need to find the pinout for the header it uses. I think the header on Jetsons (perhaps also for the Seeed carrier board, but not certain) will be “similar”, but perhaps not an exact match to the 40-pin header in the article. The setup of the carrier board itself is tied to whether it is NVIDIA’s dev kit, versus a third party carrier board. That layout is itself someone modified by variations in the device tree. I saw this is i2s, so if you find the pin diagram for the header, and it only uses something like i2s and power, I think someone familiar with i2s could answer. If there is GPIO involved, or i2c or some other control, then there is more to it (likely a device tree change).
So I suggest finding the pin diagram for the header on the microphone array, and posting that. Note that it isn’t enough to post the diagram for an RPi because only a subset of the pins are likely used…you have to know the function of each pin which is actually used.
I don’t know how many of those pins are actually used, as it covers the entire 40-pin header as a mounting point (it might be not all pins are used). However, whoever answers should be able to help with that information. We do know i2s must be set up, and likely several GPIO.
The Seeed carrier board does use an NVIDIA module, but they use an eMMC model of module, plus their carrier board has a different device tree. Many parts of their device tree will be an exact match to your dev kit hardware, but parts will not be. On top of this, it is likely that Seeed modified the device tree further for the GPIO which that microphone array uses, so I’m thinking that pins can be wired directly without any intervening hardware if the device tree can be set up to match pin function with what the Seeed carrier board is running after adapting it to the microphone array.
I guess the first question someone from NVIDIA might be able to answer is how to enable i2s. After that I think it is a case of knowing which pin to assign what function via the jetson-io (I think setup will probably be using content in “/opt/nvidia/jetson-io/”, but am not certain)