A flatbed scanner is part of a Canon Pixma MX340 multi-function inkjet, and I've freed it from its associated hinges and damper. Once freed it was relatively straightforward to remove all visible fasteners and work around the perimeter to pop loose all remaining clips. I lifted the glass top and found it surprisingly empty inside.

During normal operation I could see the scanning head and two ribbon cables through the glass. I had expected to find more components hidden along the sides out of sight. At least a switch or sensor for the scanning head to find its home position. But there's nothing, just structural ribbing and empty space.

One ribbon cable led to an assembly with visible gears, so it's probably the motion control cable.

There weren't anything else holding the scanning head in place, so I could flip it over to confirm four wires consistent with a stepper motor. The curious part is I counted five conductors in the long white ribbon cable. Either I miscounted or there's an extra wire I lost track of, for purpose I have yet to determine.

I count twelve conductors inside the other ribbon cable, leading to one end of the scanning head assembly. I guess image data comes across this wire. While the ribbon cable and its associated connector are too small for me to work with, I see at least five through-hole pins on a circuit board and I could work with that. It's something to look into more detail later.

The other end of the scanning head has a spring to help keep the imaging hardware tight against the bottom of the glass surface.

Speaking of which, I was surprised to find two separate pieces of glass. One for the large whole-page scanning window, and a separate narrower piece works with the automatic document feeder. I had expected a single piece of glass spanning across those windows. Why did Canon engineers decide two separate pieces were better than a single piece? There must be a good reason for increasing parts count and assembly complexity. Do these two pieces have different optical characteristics? Or maybe it's a supply chain volume thing. The large piece would be a high-volume piece usable on all printers and scanners, whereas the small window glass is lower volume only applicable to machines equipped with an automatic document feeder.

The glass pieces were about 3.45mm thick, far thicker than I had thought they were. I guess I got too used to LCD glass that are less than a millimeter thick and easily cracked. I've always been scared to accidentally crack the glass on a scanner bed, thinking they were just as thin and fragile as display glass. Now I know better. They're still glass so I should still treat them with care, but I won't be afraid to breathe on them anymore. Another feature I appreciated is that their corners have been beveled so I'm less likely to cut myself open on these edges.

Seeing how robust they are, I'm inclined to remove these pieces of glass for reuse elsewhere. They seem to be held in place by double-sided tape. Peeling off the tape and cleaning any remaining residue should be trivial for a glass surface.

While looking over the glass, I noticed this distinct pattern of three black stripes on a white background hiding underneath the bar between the two visible scanning windows. I think this explains the lack of a physical homing switch. Rather than adding hardware for a homing switch, the imaging sensor (which needs to be on a scanner anyway) is used to look for this pattern indicating home position. I wonder if I can spoof it by printing this pattern on a sheet of paper? That's a potential experiment for later. Right now, because I wanted to keep the entire machine functional, I had to put the scanning module back together. That's the easiest way to keep this stripe in place and visible to the scanning head while I go off and play with the rest of the device.

This teardown ran far longer than I originally thought it would. Click here for the starting point.