Let's start with a few pictures to clarify things.
Here's my test setup. New display paralleled with old, pin for pin.
The top view of the board.
The bottom view of the board.
And finally, a crazy board "feature".
This board does work, but the layout seems really crude, even for 20 years ago!
The following chips are on the board.
PIC17C756A-33/L MCU 8-bit PIC17 PIC RISC 32KB EPROM 5V 68-Pin PLCC
MM74HC4053N Triple 2-channel analog multiplexer/demultiplexer
LM324 QUAD OPAMP
2N3904 NPN TRANSISTOR
24LC128 EEPROM
DM74S00N QUAD 2-INPUT NAND GATE
Trying to trace the tracks on the board is nearly impossible with the LCD soldered in place and the crazy track routing. Fortunately, there's not much to trace as far as my issues are concerned.
Pin 3 Vo – tied to ground.
Pin 15 LED+ – tied to +5v
Pin 16 LED– – tied to ground via 10 ohm resistor. The 10 ohm resistor is verified as 10 ohms.
The replacement display I may use operates at a "horrifying" typical current of 240mA with typical al Vf being 4.2v. A 10 ohm resistor would yield 80mA, a third of the typical current. A 3.3 ohm resistor would yield 240mA, with the LEDs dissipating just over a watt.
The power supply for this board is buried in the bowels of the machine. I don't know if it would supply 3× the current.