Dear Kiran,

thank you very much for this valuable information!!
I always had the suspect that these RTC chips are little bit beasty. I have often heard about programming failures, and that these can be cured with correct power supply decoupling. But that extra pulses are produced when contacting RTC-chip, is something new for me, and probably explains some strange performance of such chips I were faced to in the past.
Such information is extremely valuable, because you only get it when being in extra close contact with manufacturer. You will never read this in a datasheet. So, thank you again.

Perhaps, may I add some explaination for this strange performance and the demand of accurate power supply decoupling?
RTC-chips are fabricated in a technology saving an extremely amount of supply current. This means everything inside is rather high-omic, e.g. output impedances of several internal driver stages. Then, there is only some non-avoidable stray capacitance necessary to have some disastrous crosscoupling! Means, when RTC-chip is hit by some RD#- or WR#-signal from mcu very fast edges of this signals are enough to cause disastrous charge injection resulting in an extra pulse. And think also about datalines, all the time having contact to RTC-chip, and all the time being busy....

In the past I preferred to interface RTC-chip with serial communication, e.g. I2C, SPI, or selfmade. May be, that I now know why... {grin}

I2C-bus is a 'slow' speed interface, using decreased slew-rate, and thereby helping in drastically reducing EMI. Also, some RC low pass filter may be inserted when directly interfacing RTC-chip to port lines of mcu: Only slightly rounding edges of square wave.

Regards,
Kai