Connecting arduino to COG LCD which is already hotbar soldered to a PCB

I want to use the LCD display from a Casio fx-85GT X classwiz calculator as I think it will be a fun and challenging project to figure out the display protocol and connect the display to a much more capable processor on a custom PCB; then I could program a calculator with more interesting functions and have control over the firmware.

I need a way to connect the 0.6 mm pitch roughly 35 pin surface mount pads (shown in the attached photo) to some pin headers so I can control the display with my arduino or teensy 4.0 through a voltage divider.

So far I have tried using an old elastomeric connector I scavenged out of an energy monitor to connect it to pads on a breakout PCB with the same pitch, but the connection has a very high resistance when there is continuity at all.

What other options are there that I could use aside trying to find and buy a COG LCD with the same dimensions instead?

Thank you in advance for your ideas and suggestions.

What is a COG?

Sounds like an interesting learning project, but you will find that there is no "protocol" to the display. The reason there are so many pins is because many of the discrete elements on the LCD have individual connections.

Unless you are adept with repairing SMD components, you are going to have a tough time making a reliable connection.

"Chip-On-Glass". I don't know if that's what the OP actually has.

(Attached image)This another one of the same display, but from a much more broken fx 85gt x calculator - you can see it has the controller directly on the glass, so I shouldn’t have to worry about driving the individual rows and columns.

I’m trying to find a way I can first observe the signals the calculator’s processor sends(I have made a basic logic analyser with my teensy4 to do that) and the replicate these signals to talk to that driver chip to control it directly.

I have so far figured out that the traces on the right (of the previous image)are the ones for data as when I short them to ground the display loses columns, shifts the pixels around or messes with the contrast. The traces on the left are for the voltage booster circuit that the driver uses to actually power the LCD elements - with 9 volts between one of them and ground, the currently active pixels are turned blue as I guess they are being over-driven.

My main focus at the moment is a way to reliably connect to those data pins to read those signals and in future send my own signals to the driver.

Thanks in advance for help.