Piezoelectric Drop-on-Demand Inkjet

A continuous inkjet is always firing and throws away the drops it does not want. It buys reliability with a gutter, a pump, a filter and a solvent that evaporates. Drop-on-demand asks the opposite question: can a nozzle sit still and produce exactly one drop when told to? Steven Zoltan at Clevite and Kyser and Sears at Silonics both answered yes in 1972, with the same component in two arrangements. Zoltan wrapped a tube of piezoelectric ceramic around a glass capillary so it squeezed inward; Kyser and Sears put a piezo diaphragm on the wall of a small chamber so it flexed. Either way, applying a voltage changes the chamber's volume slightly and a drop leaves the nozzle. Slightly is the word. Work backwards from a thirty-picolitre drop and a one-square-millimetre diaphragm has to move about ninety nanometres - which at four hundred picometres per volt is a couple of hundred volts - which is what the 1972 machines ran on - or a tenth of that once you bond the ceramic to a passive plate so it curls instead of stretching. The whole drive voltage of every piezo head ever built falls out of that one sum. What is not obvious is that the chamber rings. A pressure wave runs to the far end and back in a few microseconds, so if the next pulse arrives while the ink is still sloshing the drop leaves at the wrong speed and the wrong size. The drive is therefore never a pulse - it is a pull, a push and a damping segment, each timed against that acoustic round trip. Tuning those three numbers is most of what a print head manufacturer actually sells. And because nothing in the chamber gets hot, the ink only has to flow. UV-curable monomer, solvent, hot wax, conductive silver, live cells in buffer - all of it goes through a piezo head, and none of it through a thermal one.