"Techniques for 92% efficient LCD illumination: Waste not, want not..." 44 pages.
This app note is Part 3 in the grand saga of cold-cathode fluorescent lamps (CCFL). Part 1 was Figure 36 back in App Note 45, and Part 2 was App Note 49. This app note begins with an unbelievable quote: "One notable aspect of [the publication of App Note 49] is that it generated more response than all previous LTC applications notes combined." It is hard to imagine that this topic is so much more popular than all his other (prolific) work. On the other hand, reading through this app note, you get the sense that CCFLs are poorly understood, and that Jim's efforts were perhaps the first attempt to truly explore this under-appreciated topic.
This app note is effectively a (significant) update to App Note 49. "The partial repetition is a small penalty compared to the benefits of text flow, completeness and time efficient communication." It is dated one year after App Note 49 (August 1992 to August 1993), so it represents a considerable amount of time and effort. "Getting the lamp to light is just the beginning!"
Figures 1 through 4 show some of the additional investigation and research done since App Note 49, summarizing some typical characteristics of CCFLs. The schematic in Figure 6 is mostly copied from Figure 2 in App Note 49 (with new transistors and a new value for C1, which increase the efficiency from 82% to 88%), however, Figures 8 and 9 (with 91% and 92% efficiency, respectively) are new. Low voltage operation (shown in Figure 10) is also new. Figures 11, 12, and 13 are copied from App Note 49, while the more-in-depth discussion of LCD bias (with Figures 14 and 15) is new.
A considerable discussion of mechanical layout begins on page 11 with Figure 16. "Poor layout can easily degrade efficiency by 25%, and higher layout induced losses have been observed."
Feedback-loop stability is discussed starting on page 13. Figures 20 through 26 show some of the problems and cures thereof in the feedback loop. (Footnote 10, "The high priests of feedback..." seems to be mocking me.) The compensation approach used in these circuits is very conservative (see the 2-uF capacitors on the VC pins in Figures 6, 8, 9, and 10). "Isn't a day of loop and layout optimization worth a field recall?" (Is experience speaking here?)
The final two sections discuss dimming ("extending illumination range") and synchronization. Figures 28 and 29 improve the dimming capability of the lamp by using isolated drives (a 40:1 dimming ratio is achieved), maintaining high effeciency. The symmetric electric field around the lamp reduces the occurrence of the "thermometer" effect. Figures 30 through 36 discuss synchronization (a topic that he originally discussed, and dismissed, in Appendix A of App Note 29). "In particular, pen based computers may be especially sensitive to asynchronous components." Figures 32 and 34 abandon the Royer base drive for a flip-flop-based scheme. This approach reduces jitter at the expense of efficiency.
The best quote is the caption of Figure 20: "Destructive high voltage overshoot and ring-off due to poor loop compensation. Transformer failure and field recall are nearly certain. Job loss may also occur."
I'll discuss the appendices next time.