We present the development and the characterization of a photoacoustic NO2 trace gas detection system. The system is based on the implementation of low-cost components, i.e. a mass-produced blue diode laser and a standard MEMS microphone which is commonly built into smartphones, for example. An optimized cell design was realized by means of 3D printing. The linearity of the photoacoustic signal dependency on the analyte concentration was verified from 200 ppbV to 100 ppmV NO2. The detection limit (1σ) was determined to 33 pptV and the normalized noise equivalent absorption coefficient was calculated to 7.0 ∙ 10−10 W cm−1 Hz−1/2. The dynamic range of the system was verified to be linear over three magnitudes of order and the sensitivity was calculated to 814 μV/ppmV. The system was characterized in view of optimal operating parameters, i.e. lock-in time constant τLIA and total mass flow rate, optical performance and signal stability. The mass flow dependend response time of the system was specified to 19 s and an idealized step response to a quasi-Heaviside step function was quantified as a function of τLIA. The quality factor of acoustic resonance was determined to 21.9 and an empirical expression regarding acoustic node shifting is provided. The expression takes into account the radius of the resonator pipe and the radius of the hole, which was drilled into the pipe for microphone coupling. Furthermore, we studied the cross-sensitivity of the photoacoustic signal towards H2O and CO2, respectively.