FYI, I just repaired two Pioneer keyboards (a 17B and a 42S) using slightly different techniques, and have thus resurrected two failing units.
On the 42s, after removing the PCB, I cut off the heads of the heat stakes bonding the metal backing plate to the plastic front. (I saved the cut-off heads for later.)
A four-piece sandwich of two flat sheets of mylar with holes, one flexible printed contact circuit, and a dome sheeet came out easily. On this unit, several of the domes' internal conductive ink spots had been worn away, and a loose flake was shorting the on/off contacts.
I ordered a conductive ink pen -- one specifically advertised as flexible and durable -- and "drew" a horizontal swath across the dot in each dome. (The contact fingers are interlaced and run vertically, so I figured a horizontal line of conducting ink in the domes should do the trick.)
I reassembled everything how it seemed it should go together, and put the whole assembly in a jig that applies pressure to the LCD and keyboard contacts without re-twisting the metal tabs that normally hold the keyboard and PCB together. (Thanks to Randy Sloyer for sharing pictures of his similar jig -- such a device is absolutely indespensible for working with Pioneers.)
Nothing. Turns out, I'd not put the mylar sheets together correctly -- one of the flat sheets with holes is meant to insulate the contact flex circuit from the metal backing plate. After getting that fixed, everything checked out perfect!
To re-bond the backing plate and plastic front, I used a small soldering iron with a pointed tip. Without the PCB in place (you probably would have guessed that), I picked up a cut-off rivet head with the hot tip of the soldering iron, placed that over one of the cut-off heat stakes, then pushed the iron point through the rivet head and into the remainder of the stake. This melts the two pieces of plastic, and forces a bit more out of the stake and up toward the metal plate. I then rotate the soldering iron, remove it, and press down on the warm (HOT!) plastic with my thumb. This re-forms a sort of rivet head that, while not as strong as the original, seems strong enough.
With all of my ersatz rivet heads rebuilt, the keyboard seems nearly as good as new. It certainly functions like a new unit, though it may not stand up to abuse quite as well.
To enhance the process, I tried a slightly different tack on the 17B. I took a short length of ~ 1/8" brass tubing and rotated it while dragging the end along a file at an acute angle, creating a tiny "hole saw". I chose the tubing so that its inside diameter was just greater than the o.d. of the heat stakes.
With this tool chucked in a nice pin vise, I tried to "saw" away the mushroomed parts of the 17B's heat stakes. This was a rather more involved process than simply cutting them off, but it left more of the heat stake protruding and made later re-fastening much simpler.
The 17B had a contaminated "STO" key, and after cleaning everything up, I could simply reassemble it. (There was no need for more conductive ink.) Like I say, disassembly was much more involved, but the reassembly went like a charm, and I didn't have to melt cut-off parts onto the heat stakes to make them work.
Both of these calcs work great, and I thought I'd share my experience in case any of you are facing similar problems.
Sometime (soon?) I'll rewrite this as an article for the Articles Forum -- with some pictures.
Good luck with your own Pioneer repairs -- and don't give up on that flaky 42S!