The origin of hard X-rays (>1 keV) in polars is considered to be the post-shock region which is located at few hundred kilometers above the white dwarf (WD) surface whereas the soft photons (< 1 keV) are often assumed to be coming from the surface of the white dwarf.
This picture suggests that there should be a strong correlation among the hard X-ray photons in different energy bands. Using the RXTE data for the source AM Herculis, we found that this scenario is always not true.
We found that in three observations, the soft (2-4 keV) and hard (9-20 keV) photons were anti-correlated and in one observation the hard photons were lagging by ~190 s to the soft photons. Similar kind of delay was observed only in one another source, asynchronous polar BY Cam (Ramsay & Cooper 2002) but the selected energy band were different.
We also found the spectra pivots during the detected lag which suggest that the there is a change in the intrinsic physical/radiative processes.
We speculate that the delay is due to the bounced back material from the WD surface to post shock region.