Fix FPE in HGTD_TimingResolution where division 0/0 occured.
I know that comparing with 0 is not ideal, but in this case it seems that there's no pulse so the values stay at the defaults so at least in the q-test things look fine now. Maybe @sroe would fix it differently.