*UPDATE*: After a lot of tradeoff analysis, simulations, etc I have decided to use an H-bridge to create the triangular waveform. Segments will be controlled by opto-triacs.
An electroluminescent display is basically a "light emitting capacitor". *Why an H-bridge* ?? Electroluminescent displays require an AC supply around 1 kHz. Lower frequencies, such as 400Hz, will work but require more voltage, which puts dielectric stress on the display. Higher frequencies can reduce the voltage for a desired brightness. This is in line with their capacitor-like construction. I could have used a sinusoidal oscillator and driven a power transformer, but I've already done 2 previous designs with high-frequency step-up and ran into a lot of problems on the bench because it wasn't possible to accurately simulate the design under all load conditions. The H-bridge is well-behaved in simulations, and they match reality pretty well. This is a non-symmetric H-bridge: The "pullup" parts are just a simple on/off switch (NPN) transistor, while the pulldown parts have variable strength. Due to the high operating voltage, NPN devices had to be used because there aren't any readily-available PNPs rated above 700V that also have adequate safe-operating area (SOA) characteristics. Even finding the NPNs was a bit of a hunt. MOSFETs at that voltage have too much leakage current. *Why a triangular wave* ?? Applying a triangular wave to an electroluminescent display results in constant current. The H-bridge has 7-levels of binary-weighted programmable pulldown current (16uA to 2mA per bit), which will drive up to 8 IEL-0-IV displays. There are options to drive up to 40mA to handle up to 6 of the large I-195 displays. It's actually easier to generate a triangular wave than a sinusoid. The current needs to be adjusted based on how many segments are driven (load capacitance). *Why opto-triacs* ?? I need bidirectional current control, so that rules-out simple transistor/MOSFET control. Traditional triacs will work, but because of the way the H-bridge is built, the driver for the opto-triac is much simpler than for a traditional gate-driven triac. So, the design is being captured in KiCad and simulated in SPICE (analog) and Verilog (digital) as I inch forward. On Wednesday, March 11, 2026 at 1:55:43 PM UTC-7 Mac Doktor wrote: > On Mar 11, 2026, at 4:50 PM, Sture Nystrom <[email protected]> wrote: > > As fair I can remember the lifetime is not so long on a EL display. > > > IIRC, like many technologies EL phosphors have a half-life related > directly to how hard they’re driven but compared to others it’s really > short. > > > Terry Bowman, KA4HJH > "The Mac Doctor" > > https://www.astarcloseup.com > > "If only you could see what I've seen with your eyes."—Roy Batty, *Blade > Runner* > > -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/9c249f67-4272-43c9-9b3e-d4e8320e1595n%40googlegroups.com.
