Charles has a very good point. The CS1 pin of this part must be handled correctly. There are a couple of additional points I would like to elaborate on, if I may.

The CS1 signal is not really out of the picture after power on. This signal must be a high level voltage to access the part for any normal access after powerup.

The voltage detector circuit used to drive CS1 low when the normal +5V is not within specified operating level can be derived from a supervisor chip. You could use a supervisor designed to produce a low level reset pulse and connect that to the CS1 pin such that the pin goes high only after the power supply has stabilized.

The data sheet for the part (kudos to Charles for linking it) has a caution that the RTC chip should not be accessed for a I/O while the CS1 pin is changing. Depending upon the relative timing of the CS1 pin going high and microcontroller firmware attempting to read the clock the first time it may be necessary to insert a suitable delay to ensure this condition.

Lastly, something not covered exhaustively in the data sheet, is that it is really important to have a capacitor on the VCC pin of the RTC chip. Particularly if you have a non-rechargable battery with a series diode feeding into the VCC pin for battery backup. It is possible that there can be switching transients when the +5V goes ON or OFF. This transient can be detrimental to the proper continued operation of the RTC chip if the voltage on the VCC pin were to dip below 2 volts for even a short period of time. Placement of a good quality capacitor between the RTC VCC pin and GND can help prevent this problem. Note however it is not a surefire fix because much also depends on the type of +5 volt switching circuit that you have present to enable the +5V to the RTC VCC pin during normal operation. It is a good idea to study what happens on the RTC VCC pin during power up and power down using a digital storage type oscilloscope (DSO) to make sure that there are no transients there.

MICHAEL KARAS