Hi team,
My customer is using TPS65910 and they have found that the waveform on the passive crystals. Could you please help to explain why?
The waveform in below are: OUT of crystals, IN of crystals,CLK32KOUT(PIN38 of TPS65910)
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
Hi team,
My customer is using TPS65910 and they have found that the waveform on the passive crystals. Could you please help to explain why?
The waveform in below are: OUT of crystals, IN of crystals,CLK32KOUT(PIN38 of TPS65910)
James Zhan,
What is your question?
These all look like similar waveforms with the same frequency: ~32.5kHz, with the last of the waveforms looking like a clean, buffered clock signal.
Thanks,
Brian
James,
The maximum Coscin, Coscout value is 25pF.
The minimum Coscin, Coscout value is 12pF, as shown on page 13 of the TPS65910 datasheet.
Coscout = Coscin = CL, or the load capacitance on the crystal. You can modify the load capacitance as long as you keep the two values equal, but you have a nice clean waveform so I do not see any reason to change the capacitance.
The Output (at OSC32KOUT) waveform you are seeing is generated by the PMIC, as you can see in Figure 6-3 on page 52 of the TPS65910 datasheet. The OSC32KOUT pin voltage is generated by a current source and a switch (FET). Since the switch is referenced to GND, I see no way that the OSC32KOUT pin can generate a voltage below GND potential. This is why it is clipping at the bottom.
But the OSC32KIN pin is the input to the comparator, which generates the "32 kHz to digital block" signal. This signal is buffered and becomes CLK32KOUT, which is used by the system.
Since OSC32KIN is a very clean sine wave without clipping, and CLK32KOUT is a square wave with perfect frequency, it doesn't not matter how non-ideal OSC32KOUT is. This pin is simply used as the stimuli to start the crystal oscillator and keep it running.
Thanks,
Brian