Richard Prosser wrote:

> 8 pins on a keypad suggests a direct row x column matrix (i.e. 7 used)
> rather than BCD (which would only need 5 or possibly 6 = 4 data  + 2
> power).

  The first will be correct.  It's obviously a matrix membrane keypad.
Because a mambrane keypad brings the circuit close to your fingers, it
would be very prone to static discharge causing false actuation or frank
damage.  I *suspect* you will find the eighth connection (4 x 3 = 12
keys) to be a shield between finger and keys.

  *Old* telephone keypads had 4 by three and a common which contacted
both, but that's uncommon now.

> Can you see any logic ICs etc. anywhere?

  It'll be very thin, possibly adhesive-backed to attach to the front of
a case whilst the connector passes through a slot.  No encoding, no ICs.

> What about multimeter readings between pins when a pad is pushed - do
> you get any change? - Use a high resistance range to limit current to
> as low a level as possible - just in case.

  Unless you use an automotive lamp box, your test equipment is most
unlikely to injure the thing.  Most multimeters pass, let's see, 200
ohms = 200mV so a current of 1 milliamp.  Maybe more on diode test,
perhaps 2 or even 5.  (Don't leave your multimeter set to diode test
and connected for too long if you want to conserve the battery!)
--
  Cheers,
        Paul B.