RTC85063AT behaviour during inconsistent supply to RTC chip

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

RTC85063AT behaviour during inconsistent supply to RTC chip

970 Views
MN123_123
Contributor I

Hello!

I am trying to understand the behaviour of PCF85063A during inconsistent power supply.
In our aplication, there is no battery backup to RTC chip.The whole device is battery powered.
And we are facing issue that RTC is lagging by days and the battery terminals are found to be corroded to some extent and we are trying to find the root cause.

As per datasheet, supply voltage range for RTC chip is 1.8V to 5.5V when interface is active. When interface is inactive, it is 0.9V to 5.5V

MN123_123_0-1662453665704.png

So does this mean, when interface is inactive, RTC will continue to work/increment between 0.9V to 5.5V range.

And what are the cases when RTC will stop working/incrementing?

Thanks in Advance.

0 Kudos
Reply
5 Replies

964 Views
JozefKozon
NXP TechSupport
NXP TechSupport

Hello Mamata,

the PCA85063A will work with VDD voltage as low as 0.9V, but you won't be able to use the I2C interface if the VDD voltage is below 1.8V. The PCA85063A will stop to increment the time registers if for any reason the VDD voltage drops below 0.9V, even for very short time. When the oscillator stops an OS flag is set in the Seconds register (address 04h). This can be checked, but the VDD must be 1.8V or higher, so the I2C interface will work.

JozefKozon_0-1662462971566.png

If the VDD drops below the minimum value (0.9V), the VDD voltage must be 0V, before the power-up, otherwise the registers might get corrupted. 

JozefKozon_2-1662463301949.png

JozefKozon_1-1662463246380.png

Please make sure, that the supply voltage is stable as possible. With the changing VDD voltage, the frequency of the oscillator also changes. Please refer to the Figure 27 and Table 40.

JozefKozon_3-1662463429810.png

JozefKozon_4-1662463498684.png

With Best Regards,

Jozef

0 Kudos
Reply

957 Views
MN123_123
Contributor I

Thanks Jozef for your quick reply.

I have few questions.

1. Below 0.9V, RTC stops incremneting. But what if voltage is let's say 0.5V and then it rises to a voltage greater than 0.9V, then will it start incrementing? and will it lose initialization for selected load capacitor setting?

2. Under what conditions OS bit is set? Will it get set due to power fluctuation?

3. We are using 20ppm crystal. Manufacturer -EPSONManufacture Part Number -FC-135 32.768KA-AG0. Load capacitor value is set to 12.5pF. We are getting drift of about 4 to 5 seconds per day. So, is it as expected?

Thanks in advance

Mamata

0 Kudos
Reply

952 Views
JozefKozon
NXP TechSupport
NXP TechSupport

Hi Mamata,

1. Below 0.9V, RTC stops incrementing. But what if voltage is let's say 0.5V and then it rises to a voltage greater than 0.9V, then will it start incrementing? and will it lose initialization for selected load capacitor setting?

[A] If the voltage drops below the 0.9V, let's say to 0.5V and then it rises again, the oscillator has stopped and needs from 200ms to 2s to start again. 

JozefKozon_6-1662548733800.png

The register might and probably will get corrupted. So will not know what values are in the registers.

Yes the initialization will be also lost.

JozefKozon_8-1662549070815.png

So after the VDD drops below the 0.9V, but not to 0V, it is recommended to disconnect the VDD supply voltage, or if not possible software reset must be conducted. But this software reset requires VDD minimally 1.8V, so the I2C interface will be active.

2. Under what conditions OS bit is set? Will it get set due to power fluctuation?

[A] The oscillator can be stopped for example by connecting one of the oscillator pins OSCI or OSCO to
ground. Or if the VDD voltage drops below the 0.9V. Yes the power fluctuations might also stop the oscillator and it takes from 200ms to 2s to start the oscillator again.

3. We are using 20ppm crystal. Manufacturer -EPSON Manufacture Part Number -FC-135 32.768KA-AG0. Load capacitor value is set to 12.5pF. We are getting drift of about 4 to 5 seconds per day. So, is it as expected?

[A] The FC-135 32.768KA-AG0 crystal is not suitable to be used with the PCA85063A, because the crystal requires 9pF load capacitance, which cannot be set on the PCA85063A. This is the reason for such a high deviation. 4 to 5 seconds per day is too large deviation.

JozefKozon_2-1662548067235.png

Please choose a crystal which requires either 7pF or 12.5pF and set the CAP_SEL bit accordingly.

JozefKozon_3-1662548146596.png

 

JozefKozon_0-1662548032587.png

In our EVB a FC-135 32.7680KA-AG3 crystal is used. Please consider to use this crystal.

JozefKozon_5-1662548305301.png

JozefKozon_4-1662548274287.png

https://www.digikey.com/en/products/detail/epson/FC-135-32-7680KA-AG3/5259897

With Best Regards,

Jozef

0 Kudos
Reply

950 Views
MN123_123
Contributor I

Hello Jozef,

Thanks for your prompt and detailed reply. We will look into using the crystal suggested by you.

 

Mamata.

0 Kudos
Reply

945 Views
JozefKozon
NXP TechSupport
NXP TechSupport

Hello Mamata,

you welcome.

With Best Regards,

Jozef

0 Kudos
Reply