This letter investigates the performance of affine frequency division multiplexing (AFDM) under finite-resolution analog-to-digital converters (ADCs) and practical hardware impairments, as well as realistic channel estimation. The additive quantization noise model (AQNM) is adopted to characterize ADC quantization errors, while the generalized complex exponential basis expansion model (GCE-BEM) is employed to model time-varying channels. Based on these models, we derive a closed-form expression for the normalized mean-square error (NMSE) of the channel estimation in the presence of ADC quantization and hardware impairments. Furthermore, we theoretically analyze the impact of finite-precision quantization and non-ideal hardware conditions on the bit error rate (BER), revealing the intrinsic relationship among BER, quantization resolution, and hardware impairments, thereby offering useful guidelines for practical system design. Simulation results validate the derived NMSE and BER analyses and demonstrate the sensitivity of AFDM performance to ADC resolution and hardware impairments. Additionally, our results confirm a resolution saturation region of AFDM around 6 bits, at which point thermal noise dominates and quantization can be disregarded.