Sharp PC-1600 keyboard issues

[Josh Dersch]

Another day, another project (the DAP will wait until I have the wiring 
to support it...).  This week a bunch of "Refurbished" Sharp PC-1600s 
showed up on eBay for cheap and I snagged one; it arrived this evening.  
It looks like the unit is effectively new-old-stock; it doesn't seem to 
have been used since Sharp refurbished it back in the day -- it arrived 
with the original Sharp-branded batteries in the compartment.  
Amazingly, only one of the four batteries leaked, but only a tiny bit 
and there's no obvious corrosion anywhere from it (whew).

The PC-1600 powers on and mostly behaves normally after an "All Reset" 
but the keyboard is having trouble; here are the symptoms:

- Keys A-K on the 2nd row of the QWERTY section do not function ('L' does).
- The "Shift" key doesn't work properly -- hitting it doesn't cause the 
"Shift" annunciator on the LCD to light up.  However, it does seem to do 
*something* in that if you hit Shift+another (working) key, nothing 
happens at all (for example, Shift+5 should result in a left curly brace 
but does nothing at all.).
- The "SML" key does not put the machine in lowercase mode, instead one 
of the Kanji (i believe) annunciators gets toggled. This has no other 
obvious effect (uppercase letters are still produced from the working 
letter keys...)
- Hitting the "KB II/Click" key (farthest right under the LCD) turns 
"Key Click" mode on even though Shift is not being depressed.  (Or maybe 
this is normal behavior -- I can't find this key described in the manual...)

Otherwise, the thing seems to be working OK, it'll calculate fine and I 
can enter/run small programs (that don't use the home row for entry...)

I've found the Technical Drawing set here: 
http://www.sharp-pc-1600.de/Schaltpl%8ane%20A3/Technical_Drawing_DINA3.pdf 
which makes it obvious that the dead home-row keys are related to the 
KIN3 line from the keyboard matrix; I note that the keyboard signals are 
brought to the edge of the PCB and connected to a membrane layer which 
contains additional circuitry, including what looks to be an IC that 
does the keyboard decoding.  My first guess would be dirty contact 
between the PCB and this membrane (esp. since the lines in question are 
close to a hole in the battery compartment where the aforementioned 
30-year-old batteries were...) but I'm a bit nervous to disassemble and 
clean this without knowing how it's held together.  Anyone have any 
experience here? Any tips?

Thanks as always,
Josh