OS bit on PCF85263ATAJ

[SCHlabs]

Hi Everyone:

I have a project with PCF85263ATAJ chip, Real Time Clock with battery backup. The chip works almost well:

Works fine: Oscillate / The clock is accurate / Communication /

Problem:When i left turned off the device all night ( on battery mode), the clock works but the bit OS is activated. The time is accurate and actual.

I mount a setup and perform the following test:

1) power ON our PCB,

2) Set the actual time and clear the OS bit

3) Power OFF

4) wait entire night or weekend

5) at morning read with an oscilloscope that the oscillation is present at 32.768

6) Power On

7) The OS bit is 1, and the date/time is correct

The battery (CR1225) say 3.1v

Any hint will be appreciated.

[SCHlabs]

Any News?

[SCHlabs]

Any News?

[SCHlabs]

Any news?

[SCHlabs]

Dear Tomas,

The power fail is not the only reason to activate OS bit. I can force the OS bit activation touching the any of both xtal pad with the oscilloscope probe. The oscilloscope probe is a 3pF capacitor and 1 meter of shielded cable.

But I can ensure that the oscillation is not

I do the next test: I put the oscilloscope probe on the crystal pad, ( and I see the oscillation on screen) I power off, and power on again after 2 minutes, and I can watch that the oscillation never was stopped ( see the attached images)

Even I try to watch if the oscillator was stopped for more than 1ms, and the oscillator works fine in every moment...... fine but the OS bit was activated.

Best Regards

Christian

[SCHlabs]

Dear, i understand that you say. But does not work in this way. If i touch any pin of crystal with the oscilloscope probe (3pf and 1m of shielded cable) this bit will be enabled.

I put the oscilloscope, turn off and turn on, and measure on Xtal the oscillation. Searching on power on/off the oscillation never stop.  After take the attached photos i put the oscilloscope to find a glitch (more than 1ms of oscillator stop), and the oscilloscope never was triggered, so the oscillation was continued in every moment (on/off and transitions).

I am sure that Vbat never fall under 0.9v, Vdd obviously yes go from 3.3 to 0 and viceversa.

There are another situation that trigger this bit, maybe undocumented.

Best Regards

Christian

[SCHlabs]

I can force it putting the oscilloscope probe in one terminal of RTC. The OS bit is automatically activated but the time is not lost. This force to the firmware a RTC reset (and the time to zero).

We comment out the reset line, but the OS flag is still activated

[SCHlabs]

Dear, I change the PCB layout, and the problem persist. Not in all PCB, not in every time. But the clock back to zero. I was looking for a "layout app note"  but i cant find it. The schematic is attached.

[TomasVaverka]

Hi Christian,

OS flag set means that PCF85263A internal oscillator oscillation has been stopped and there are two possible reasons to cause oscillator oscillation stopped which are either POR or VDD voltage goes below the min operation voltage (0.9V). 

BR, Tomas

[SCHlabs]

Dear Tomas,

The power fail is not the only reason to activate OS bit. I can force the OS bit activation touching the any of both xtal pad with the oscilloscope probe. The oscilloscope probe is a 3pF capacitor and 1 meter of shielded cable. So imagine that the noise and capacity of the probe alter the stability for a short moment.

But I can ensure that the oscillation is not interrupted

I do the next test: I put the oscilloscope probe on the crystal pad, ( and I see the oscillation on screen) I power off, and power on again after 2 minutes, and I can watch that the oscillation never was stopped ( see the attached images)

Even I try to watch if the oscillator was stopped for more than 1ms, and the oscillator works fine in every moment...... fine but the OS bit was activated.

Best Regards

Christian

[SCHlabs]

do you know for any news?