Definitions

One definition of thermal printhead life is pulse life. Kyocera tries to guide printer manufacturers so that a pulse life of 10⁸ pulses is achieved. Pulse life is a function of maximum heater temperature. High temperatures shorten the life of thermal printheads and increasing the power or pulse width will increase the peak heater temperature. The amount of cooling time between successive pulses reduces the peak heater temperature, so printer speed (cycle time) also determines pulse life. The graphs on this web page show the maximum recommended power as a function of pulse width, for different printer speeds.

Power is normalized to power density to make these charts applicable to printheads of different heater sizes. A maximum power density is valid for the small range of heater area sizes used either in 200 dpi or in 300 dpi printheads. Plugging a 200 dpi heater size area into both charts would quickly show that the power density normalization is not valid for the big heater area jump between 200 dpi and 300 dpi. Similarly, different graphs would be applicable to other dot densities. The maximum energy rating given in an applicable printhead specification represents one point on one of these charts. These charts could be used, with caution, to determine other acceptable operating points. These charts are only valid for Kyocera's current most popular heater material as used for fax, POS, bar code, etc. Other charts are needed for the heater material in some super-flat resistance printheads or for the heater material in certain old printhead models. Other charts are needed for thin glaze printheads or for edge-type printheads.

These Maximum Operating Condition Charts (MOCC) are determined from actually testing heaters until they fail, by gradually increasing the power level. The voltage is applied by probe directly to the heater, so driver IC losses and limitations do not affect the observed maximum power. Then a safety factor is applied, meaning that it is OK to operate printheads at the maximum power density. The safety factor assumes that the printhead is in good contact with the media. If operation in air is not prevented, then the maximum power should be limited to 80% of what is shown on the charts.

For long printhead life the print energy should be decreased 1% per degree Celsius over 25°C, as measured by the heat sink thermistor. For consistent print quality with most media, print energy should be decreased even more, which is that much better for printhead life. On the cold side of 25°C, Kyocera allows a maximum 1% print energy increase per degree below 25°C, down to the ambient operating temperature limit of 0°C. Cold ambient operation is more troublesome than hot ambient. A well-controlled external printhead heater will improve both print quality and printhead pulse life at cold ambient temperatures.

Compensating for Differences in Printhead Average Resistance

Each printhead has an average resistance (R_AV) that is the arithmetic mean of the resistances of all the heater elements on that printhead. It is typically printed on the label on the printhead. The average resistances of actual printheads vary significantly about the nominal average resistance for that model printhead. At a fixed heater voltage, average power will have the same range of variation as resistance, because P=V² /R_AV. If all printers used the same pulse width (T_ON), then the print energy would vary by the same range because Energy = R_AV x T_ON. This uncompensated variation in print energy would cause a noticeable variation in print image quality and an unacceptably short life for printheads at the low end of the average resistance range.

The simplest way to compensate for variable printhead resistance is to use a variable voltage power supply and adjust the voltage (V_H) at printhead installation so that the print power is constant. This print power, expressed as power density, would be a horizontal line on the MOCC and maximum pulse width could be easily read from the MOCC.

More typically, printers use a fixed voltage power supply. Therefore print power becomes a range on the MOCC. The following example is for a KPA-80-8MPA1 printhead. Its heater area is 0.11mm x 0.132mm = 0.01452mm². The specified average resistance is 660 ohm +/- 15%. The typical operating voltage (V_H) is 24V. The voltage loss (V_L) within the driver IC is specified as 0.9V, which for simplicity is always specified as a constant because it is small compared to V_H. The net power that generates heat in the heater element heater, (V_H-V_L)²/R_AV is determined by the voltage drop across the heater element (V_H-V_L), which is the same parameter that was controlled when the MOCC data was produced. The table below shows the calculation of three power density levels starting from minimum, center and high R_AV values.

The resolution of this printhead is 8 dots per mm, making it a 200 dot per inch printhead. The MOCC below duplicates the 200 dpi MOCC shown previously. On it are plotted the three power density levels as horizontal red textured lines.

The printhead specification shows pulse width as 0.26ms at 0.841w/dot applied power and it shows 0.032w/dot as the power loss in the driver IC. Therefore net power = (0.841-0.032) = 0.809 w/dot and power density = 0.809 / 0.0145 = 55.7 w/mm² . This point falls on the 1.0ms cycle time line confirming that the MOCC and the specification agree on this single point.

If a printer design does not choose to adjust pulse width as a function of average resistance, then the single pulse width for a cycle time is determined by the intersection of that cycle time line with the maximum power level line. The minimum available print energy is given by the product of that single pulse width times the minimum power level.

For example, Kyocera routinely assumes that 24 mJ/mm² is the energy density required to darken typical label media. A dotted line marked by purple filled diamonds shows this constant value on the above MOCC. The pulse width that will darken this media at all power levels is given by the intersection of the media line with the minimum power level, in this example at 0.5ms. How fast can this non-adjusted printer go while still achieving a long printhead life? The cycle time that will allow 0.5ms pulse width at maximum power happens to be visible without interpolation as the blue line (marked by "X"s) for T_CY=5.0ms, which is approximately 1 ips. This would not be acceptable in today's competitive bar code printer marketplace.

Note that the slope of the constant required media energy line is steeper than the slopes of all cycle time lines. If the printer design continuously adjusts the pulse width as a function of average resistance to give a constant print energy of 24mJ/mm², then the print speed is determined by the cycle time line that intersects the maximum power line at 24mJ/mm². This unknown cycle time is between 2.0 and 5.0 ms, perhaps about 2.8ms. The formula for calculating print speed (ips) is ips = (1000) / (T_CY * lpi). The print line density in this example is 8 lines per mm or 203 lpi. Then print speed =1.76 ips, a 76% improvement! The print image quality would most likely be better, because without adjustment, the print energy would be excessive at higher power levels.

Basis for History Control

Most MOCC have a line for 1.0ms cycle time that gives the maximum pulse width if a heater fires on every print line. If a dot printed only on every other line, then the heater would be fired every 2.0ms. A longer pulse width is allowed for 2.0ms cycle time. Similarly, if a heater had not fired in the prior two cycles, its effective cycle time would be given by the 3.0 ms cycle time line. The purpose of history control is to increase the print speed yet still allow the use of a longer pulse width when a heater is cold.

Continuing with the previous example, the maximum pulse widths for a 2.8ms cycle time are adequate to darken the media. Using one level of history control, a printer can go twice this fast which is 1.4ms. 1.4ms cycle time line is drawn as a dotted line marked by yellow filled triangles and it shows the pulse widths for continuously fired heaters. The assumption is that there will be sufficient retained heat from a previously fired dot to darken the media when a heater is fired at the lesser pulse width. This is usually the case, because over the history of thermal printer development, excessive retained heat has usually been what limited print speed.