PCF8523T/1 no communication on I2C

取消
显示结果 
显示  仅  | 搜索替代 
您的意思是: 

PCF8523T/1 no communication on I2C

2,461 次查看
boer59
Contributor I

Hi,

We see a problem in our production related to to the RTC PCF8523T/1. About 4% are failing at the first test in the end-of-line test. What the test do is simply, read time from RTC, turn off power for a number of seconds, turn on power and check how many seconds has elapsed by reading from the RTC.

I have checked more then 10 units. What is "magic" is that some of them comes to life (the RTC) and start to work as expected.

What I can see is that on units that are not working, the I2C communication is not working. There is no ACK that can be seen in the communication. See attached pictures. One is a unit working OK, the other NOK.

Can this be a problem related to Reset and the crystal not operating properly at startup ? The crystal is specified for 12,5pF and default value in the PCF8523T is 7pF.

BR

Bo

0 项奖励
回复
1 回复

2,453 次查看
JozefKozon
NXP TechSupport
NXP TechSupport

Hi Bo,

the PCF8523 has a battery backup input VBAT pin. I see you are using a supercapacitor as a backup supply. The battery switch-over circuit monitors the main power supply and switches automatically to the
backup battery when a power failure condition is detected. Accurate timekeeping is maintained even when the main power supply is interrupted.

JozefKozon_0-1624357815660.png

The reset shouldn't happened if the switch over circuit and backup system works correctly.  

JozefKozon_1-1624357952739.png

If the reset is asserted, then yes, the CAP_SEL bit in Control_1 is set to its default value 7pF.

JozefKozon_2-1624358080065.png

Is it possible for you to check the CAP_SEL bit value in the Control_1 register? Or is not possible because of non functional I2C communication? 

Before you test the RTC have you enabled the battery switch over function in the PM[2:0] bits? Could you measure voltage on the VBAT pin during your test? 

If the reset had been asserted and the CAP_SET bit in Control_1 register is set to 7pF instead of required 12.5pF there should be a change in oscillator frequency. Could you connect a pull-up resistor and to the CLKOUT pin and measure the oscillator frequency. For the 7pF the frequency should be higher, than the 32.768kHz, because it takes shorter time to charge and discharge the 7pF capacitor, than the 12.5pF capacitor.

JozefKozon_3-1624359051413.png

 

With Best Regards,

Jozef

 

0 项奖励
回复