<![CDATA[Make M144 - Audio Module true STEREO]]>

<![CDATA[Make M144 - Audio Module true STEREO]]>I found a design issue (or corner cut) in the HW design phase that dramatically reduce the versatilty of this extremely nice board.
On the Audio signal path from Input LINE/MIC 3.5mm TRS, there is a circuit that feed the microphone signal from the IN/OUT TRRS Jack.
I get the logic behind this decision but the implementation affect the frequency response on the LIN1 path.
The result?
You cannot use the board to process analog stereo signals as RIN1 frequency response is flat at the HW input and LIN1 isn't!

My solution would be to add an electronic switch (maybe a TS12A12511) on the MIC_MONO_IN between R9-C14 node and FSUSB42MUX to cut that path entirely.

The control signal can be tied with LIN_MIC_PU_EN from the STM32G030; when the Headphones Electret is energised, the path to LIN1 would be restored.
A better solution would be to make the control separate and use one of the available GPIOs on STM32G030 and a separate I2C command on 0x33 address.

It is so simple...

This will eliminate the need to physically remove C9 in order to get a flat response for BOTH Line inputs thus having a STEREO Audio Module.
See this also: My findings about this

I even asked M5 AI for this issue and it was funny because it didn't knew about this issue:

If you're seeing different frequency response between left and right, the most likely cause is:
One channel's HPF is enabled while the other is not (rolls off low frequencies)
Different PGA gain settings between channels
The two physical mic inputs may have different input impedance/coupling characteristics

Cheers,
Adrian

]]>https://community.m5stack.com/topic/8264/make-m144-audio-module-true-stereoRSS for NodeWed, 17 Jun 2026 16:08:49 GMTWed, 17 Jun 2026 03:57:50 GMT60