I'm not sure if this counts as "test equipment", "projects" or "insanity" but I figure it's worth a post

I own two Solartron 7150Plus benchtop digital multimeters. They've both had the Schaffner AC filters replaced (one after it failed spectacularly, one pre-emptively to avoid failure) and are showing their age - the display backlight LEDs are tired and dim, and the standard banana plug jacks won't take shrouded test leads.
I started refurbishing one a few months ago -- I successfully replaced the LEDs, then drilled out the panel to take some new Hirschmann shrouded 4mm sockets. Sadly my drill bit got stuck on the plastic, the G-clamp slipped and the panel went spinning. This broke (more like utterly destroyed) the Mylar flexprint cable which ran from the keypad to the front panel PCB... I tried to get a new keypad or front panel assembly, but they're like hen's teeth. You can't get them for love nor money. So I decided to do something different... I made a new keypad.
The original keypad is an AMP-made thing consisting of a black plastic base with four locating pins (held in the frame by a locking washer), a Mylar sheet with silver ink printed on it to form the conductive pattern and cable, and a white frame (made of a plastic which tends to weaken and yellow over time) which covers the front and provides the keycaps. The whole assembly is about 3mm thick, with the front white plastic accounting for most of this. My plan was to reuse the white plastic but replace the black support and Mylar section with a PCB and tactile switches.
I knew extremely low profile tactile switches existed -- I'd seen them in cheap digital cameras (like many engineers, I occasionally tear down broken-beyond-repair hardware for fun). After a few hours of searching, I found the E-Switch TL3315NF160Q in the Farnell catalogue (P/N 2103642). This is a 4.5mm square switch with a thickness of only 0.55mm and a switch travel of 0.2mm. Perfect. I also ordered a sheet of single-sided 0.8mm photoresist coated FR4 PCB material to produce the back PCB. This meant the total height of the finished unit would be comparable to the existing part.
I spent a few hours measuring the white plastic piece with my digital callipers, then created a scale drawing in LibreCAD (the open-source fork of QCAD). Once I had a CAD drawing of the keypad, I measured the positions of the actuator pins and laid out a PCB in EAGLE. I etched it and assembled it, but the actuation pins on the back of the keycaps were too large and fouled on the switches. After about an hour of trying to alter the pins to work, I ended up breaking the plastic frame... time for drastic measures.
I spent another few days designing a drop-in replacement keypad in OpenSCAD, then sent the resulting STL file out to 3DPrint-UK to have it 3D printed. What I got back was a perfect replica of the original keypad, but in SLS-printed nylon instead of the original mystery plastic (probably ABS), and with the texture of a Trebor Extra Strong mint. (I'm still trying to figure out how to add a hard-wearing glossy finish).
I combined the keypad and the PCB, soldered in a ribbon cable to attach it to the front panel board and now I can control the meter from the front panel again

The front panel, by the way, is glued on - removing it involves levering it up along the edge and using a wide putty knife to pull it away. Before you put it back on, you'll need to remove the old glue and replace it. The best way I've found of doing this is a couple of household scouring pads -- cover the front panel in masking tape to protect it, then scrub the old glue off. If it won't come off, score a crosshatch pattern in the glue with an X-acto knife.
Now I need to find a suitable glue to reattach it -- I'm thinking E6000 (aka Goop) or Evo-Stik Impact might be the best choice. The problem is, if the glue is too hard to remove, the keypad becomes irreparable (which may be necessary if the wire breaks away from the solder joints).
Nearly there, though!
