HP48SX - shift keys won't stay shifted - preliminary report

[W_Max]

I very well could be misunderstanding something, but would voltage drops only affect whether a key is detected?

I think, the reason is key press itself (some sort of leakage) + activated subroutine (rising CPU current). Other symptom is notable screen contrast cyclic reduce, while key pressed. If power regulator is faulty - drops may be critical for stable CPU functioning, causing "strange" side effects.