Errata[68] CLOCK - nRF52810

[buja]

Dear Team,

We are using the following 32 MHz crystal from Epson (datasheet attached) on our nRF52810

The datasheet does not say the start up time for the crystal and hence we do not know if it takes more than 400 us to become stable. Do you think this will cause an issue due to the errata ?

regards,FA-128_Q22FA12800452_en(1).pdf

[Vidar Berg]

Hello,

The startup time likely less than 400us, but you can try to measure it using the snippet below to be sure.

NRF_TIMER1->BITMODE = TIMER_BITMODE_BITMODE_32Bit << TIMER_BITMODE_BITMODE_Pos;
NRF_TIMER1->TASKS_CLEAR = 1;
NRF_CLOCK->EVENTS_HFCLKSTARTED = 0;
NRF_TIMER1->TASKS_START = 1;

NRF_CLOCK->TASKS_HFCLKSTART = 1;
while(NRF_CLOCK->EVENTS_HFCLKSTARTED == 0);
NRF_TIMER1->TASKS_CAPTURE[0] = 1;

NRF_LOG("HF Clock has started. Startup time: %d uS\n", NRF_TIMER1->CC[0]);

Best regards,

Vidar

[buja]

Hi Vidar,

Thank you for your reply.

I am a bit confused, The errata says "EVENTS_HFCLKSTARTED may come before HFXO is started.", which conveys that EVENTS_HFCLKSTARTED cannot be used to confirm whether the HFCLK is started. Even if EVENTS_HFCLKSTARTED is 1, there are chances that the CLOCK is not actually started (that is what errata says).

So how can you use the statement in line number 7 to check if the HFXO has started.

while(NRF_CLOCK->EVENTS_HFCLKSTARTED == 0); 

[Vidar Berg]

Hi,

Sorry, I forgot this detail about the errata. But please note that if you use Nordic SDK that rely on the 32MHz, then all protocols will always startup the external 32MHz crystal 1.5ms prior to when it's required, so even if the startup time is more than 400us this Errata 68 is not really a problem. Otherwise I think you will need to come up with a test verify to verify if the oscillator is stabilized after the EVENTS_HFCLKSTARTED event. You could, for instance, measure a PWM output signal while HFXO is being started.