Multiplexing Nixies

Jules Richardson wrote: I don't have a definitive answer, but I think the issue is more peak current than hot-cold switching. Whether the life-determinant is absolute peak current or peak current integrated over time though, I don't know. Most test equipment is non-multiplex, while most calculators use multiplex mode. In calculators the duty cycle is typically 1/12 to 1/16. The NIXIE's used in test-equipment vs calculators are, by my observation, in pretty much mutually exclusive sets: I've rarely if ever seen a NIXIE type that was used in one piece of equipment in one mode that was used in another piece of equipment in the other mode. Consequently I have wondered whether there was a difference between NIXIEs made for mux mode vs. not (different gas pressure perhaps?), and have had my eye out for manufacture data that might answer that, but haven't run across anything that actually specified multiplex mode criteria. Also, just by experience, I run into more bad NIXIEs in test equipment (non-mux mode) than in calculators (mux mode). Granted that it may well be other issues that are more significant in producing those distinctions in observation (numeral size, operation behaviour, period, etc.). The anode limiting resistor gets dropped in value for mux mode use (although not in linear proportion), with the consequent increase in peak current. One or two Sudoku rows in a multiplex set might give some benefit of multiplexing but still keep the duty cycle and peak current in 'typical' range. I'm tempted to also build in an infra-red motion detector (adapt one of those outside light sensors) to reduce un-utilised on-time in such situations. I could supply you with some display-driver schematics from calculators, for examples of how the anode drivers are typically done, if you like.