With suitable experimental measurements, it is possible to investigate the performance and quality of the MAC and physical layer implementation of different powerline communication equipment. In this paper, we present a new method for benchmarking and test of commercially available powerline adapters' ability to recovery from the influence of overload noise on the powerline network. Compared to the RFC2544 method, this method is slightly different and provides more details as to the performance of the powerline adapters both during noisy periods and after the recovery. Five different commercially available powerline adapters were tested with respect to their ability to recover from a noise influence on the powerline network. The adapters show very different performance depending on the implementation of the software and hardware in the adapters. Some modems recover rapidly to the maximum throughput level, whereas the throughput of other modems increases within adaptation recurrence. It reaches intermediate, stable throughput levels lower than the maximum throughput measured. After some time, it recovers to the maximum throughput level.