Hi all,
a few comments on the remarks of my solution.

Andy Neil wrote:
-------------------------------
Sometimes, that test will pass even when there's no chip fitted!

The problem is, if you write and then read back immediately, there could be enough capacitance around to "hold" the data and give a "pass" result.
Especially if you're on a breadboard with wire-wrapped connections!
-------------------------------
Why SHOULD you test with no RAM chip connected ??? And I assume, of course! the hardware connected PROPERLY, which is important for ALL projects!!!

phillip m gallo wrote:
-------------------------------
... you can get fooled if the part or your circuit has page addressing problems.

It is safer to ....
-------------------------------
I agree, your method is better, testing more aspects of the RAM-chip ;-)

Henk