[en] Philips 1 MW CW Klystrons have been well operated in TRISTAN for almost 7 years. Problems like degradation of gun insulation and abnormal anode emission, both caused by excess barium evaporation from cathode and by emission of electrons from hot electrodes of electron gun, have been successfully solved by the use of M-type cathode as well as of redesigned gun with decreased electrode temperatures. The electrode temperatures are, however, still yet to be optimized. Modifications should be refined much more. Positive spikes of anode current and Side Band Oscillations (SBO) which are caused by back streaming electrons from collector and/or output gap have been studied intensively. Much improvements have been made so far by taking a good combination of modified drift tubes, anode shape, number of cavities and redesigned output transitions. (author)