Secret is finally revealed, I be able to correctly drive the MN15439A. It’s turned out that the shift register is 288-bit wide, not 286-bit as I thought. Which is identically same size as MN14440A (just the vetical pixels number and grid number are different). I already made the commit to STM8 code on the STM8L sdcc template repo. And here’s the image :

There’s still 2 problems, on hardware side the Inductor get crazy hot. Physically bigger inductor with same value should help or add some thermal pad and tiny heatsink for it. On software problem is, Right now I’m implementing the driver code running on stm32 (project VFDHack32), the current code (W.I.P.) still not working and also takes a lot of memory space. I might need to rethink the code again soon.
The STM32 code is available at my GitHub repo https://github.com/TiNredmc/VFDHack32