We have encounted a couple of times a circuit where, while trying to figure out what's wrong, touching a particular point by a scope probe removes the original problem... Just it would be too expensive to supply a scope with each such device :-)

Try a small cap around 10-100pF from pin 2 to ground - it might have picked up interference from the other signals by capacitive coupling in the cable - try a 1:10 probe maybe it won't damp it down and you will see it as spikes in the clock.

But it's just a quick guess.

Jan Waclawek