Skip to content

Spectrum Analyzer Hardware

Raspberry Pi Pico 2

The original Pico will work, however the clock is slower and it does not have built-in floating point hardware and DSP instructions to accelerate the FFT algorithms.

Microphone

Low Cost Sound Sensor

sound detection sensor

You can purchase low-cost sound sensors on eBay for about $1 each. These devices come with an analog output that can be connected to the analog-to-digital GPIO of your Pico. The low-cost device I used had a VERY small signal of only a few millivolts. The ACD is looking for signals in the range of 0 to 3.3 volts. To be useful, these devices should be passed though a small amplifier to reach a voltage closer to 3 volts.

The other problem we have with these low-cost microphones is noise. The reference voltage on the Raspberry Pi Pico has noticeable nose from the other circuits drawing power.

INMP442

We like the INMP442 since it has an I2S interface and very low noise.

Display

We use a 2.42" OLED display with a SSD1306 driver.

If you're prototyping on a breadboard or using standard jumper wires, stick with 4–8 MHz to ensure signal integrity and reduce data corruption.

Breadboard

We use 1/2 size mini solderless breadboards with 400 tie-points. This is enough to place the Pico on the breadboard and it includes room for three mode setting buttons to change the mode of the Spectrum Analyzer.

Our breadboard purchasing guide can be helpful.