Yes, I was using ccx-2.15, and in both cases the work done by the damping forces in step 2 was exactly zero.
Ok. Then you should just check if the values reported by ccx 2.22 make sense. I would switch to the newest release anyway, many bugs have been fixed since 2.15.