4 posts / 0 new
Last post
silver liu
Offline
Last seen: 5 years 8 months ago
Joined: 2014-03-29 01:48
Question about P1-2 and P1-3 GPIO

Hi,
from the Application note (ANB_009), it is described that P1-2 and P1-3 are not suggested to use.

But we got problem to allocate GPIO to my project.

is it completed blocked or for some cases, it is still possible?

we are using WLCSP34.

Based on our scenario, we can use these 2 GPIO in very low frequency (simply use it to control the power switch, it is more than 2 minutes pertime in worst case.

Can we use it?

if we can use it, what we need take care of?

thanks and best regards

Silver

JE_Dialog
Offline
Last seen: 3 months 3 weeks ago
Staff
Joined: 2013-12-05 14:02
Hello Siliver, we have no

Hello Siliver, we have no test data for slow single edge events like the one you describe (turn switch on/off) - we only have for sine and square wave at frequencies > 10 KHz. The higher the voltage and the higher the frequency (upto to 500KHz), the higher the PER value.
In general, frequencies < 10 KHz at P1_2 and P1_3 are less sensitive to the 16 MHz Xtal oscillator.

We would anticipate single edge events will not disturb the 16 MHz Xtal oscillator much. If possible, the edge rate would need to be slowed down e.g. with some small amount of capitance. There is no edge rate control in the I/O of the DA14580. Overall we beleive the effect on PER would be very limited for a single edge event, but we would recommend you to check this.

Best regards, JE_Dialog

silver liu
Offline
Last seen: 5 years 8 months ago
Joined: 2014-03-29 01:48
Hi,

Hi,
thanks!
Looks still in risk!
Is there any special action(for both software and hardware) we can take to reduce the risk? like we turn off/Disable something inside the chip? and after the event done, we recover it to normal?
regards
Silver

JE_Dialog
Offline
Last seen: 3 months 3 weeks ago
Staff
Joined: 2013-12-05 14:02
Hello Silver, a disturbance

Hello Silver, a disturbance of the 16MHz clock we anticipate would not be very critical as long as you do not perform any frequency sensitive tasks  (e.g. Bluetooth or accurate PWM) - but we have not tested for any other secondary effects on the MCU, so cannot make any guarantee. There are no itnernal tips or tricks unfortuantely.. The only thing you can do to reduce risk is to slow the edge through a series resistor on the output. Best Regards, JE_Dialog