One possible way to approach this would be to write a simple function generator (square wave) and sweep it across the audio spectrum, while also recording the s

One possible way to approach this would be to write a simple function generator (square wave) and sweep it across the audio spectrum, while also recording the speaker output level by way of a connected microphone. The resonant frequency of the speaker (or the microphone) should be observable as the peak amplitude seen by the mic. Once known, just set the generator to that frequency and blast away for a while.