I have a VFD unit out of a piece of test equipment that has developed leakage between the anodes (plates). This leakage (~50k ohm to ~250k ohm) has been verified after physically removing the driver components which completely isolated the grid, heater and anodes. The only other leakage path would be the PCB as the VFD is still mounted on the PCB, however I have thoroughly cleaned (alcohol) the areas on the PCB where the display is mounted. The VFD has very low hours of use and the brightness (emission) is still strong. The symptoms of the anode leakage manifests itself as 'ghosting' segments.

Does anyone know for the mechanism for this failure? The vacuum appears OK as the getter indicates no problem. Is there a solution to remedy this issue or is the display ready for the trash?