Hello,
I am using the MIMXRT1061CVL5B MCU and I have stumbled upon a problem, in some cases after a power-down and then power-up, the micro-controller doesn't start. After some investigation, I found out that in this case where the controller doesn't start it is always the case that in power-up the power supply doesn't have a sharp rising time (it takes approx. 20 ms till it reaches the 5 V) and the oscillator doesn't oscillate even after the power supply reaches the 5 V the oscillator still doesn't start, the oscillator only starts when there is a sharp rising time in the power supply. Here you can find the schematic regarding the oscillator in my project.
Is there anything you can point out that could be causing the problem?
Thanks in advance.
Hi Diego,
thank you for your reply, we have actually solved the problem but I figured I explain what was the problem exactly in case it happens to someone else.
As you can see in the picture below, we have added the R-C circuit to implement the 1 ms delay between the DCDC_IN and the PSWITCH, the problem was, with the ramp power-on the capacitor was charging before the DCDC_IN signal comes to a stable HIGH and by the time it gets there the 1 ms delay in no longer correct (in this case it is less), we have an external reset monitoring circuit and we have connected the PSWITCH to the REST so that in case of performing a reset the capacitor would discharge, we have also increased the tau(RC) value as a back-up solution to make the charging go slower and therefore even if the power-on signal is not sharp enough the capacitor would not be charged to the fullest before the DCDC_IN signal stabilizes at HIGH. We have tested both solutions and they work perfectly.
Best regards.
Hello Diego,
thank you for your reply, sorry I have uploaded the wrong screenshot, I have already taken this recommendation into consideration and changed it a long time ago. You can see it below.
The problem is not that the oscillator takes a longer time to start, but rather that it doesn't start at all.
Thank you in advance.
More info:
- After a power-down and a ramp power-up (not sharp), the controller doesn't start, the crystal doesn't oscillate and a hardware reset doesn't restart the controller.
- I have controlled the VDD_SOC_IN signals which are connected to DCDC_SENSE (see image), when the controller is running it is 1.25V, and in this specific situation, it is totally 0V. This 0V value is the corresponding value to the SNVS mode.
Hi @Lukaz
We need to make sure that your design complies with the power-up , and power down sequence.
For example , here is an snapshot that I took the other day, from our RT1060 EVK. The first three signals from the above diagram are captured.
Do you have capture of the supply rail that rises sharply sometimes, against others?
Do you implement an external reset monitoring circuit?
Many thanks for your patience,
Diego
Hi Diego,
thank you for your reply, we have actually solved the problem but I figured I explain what was the problem exactly in case it happens to someone else.
As you can see in the picture below, we have added the R-C circuit to implement the 1 ms delay between the DCDC_IN and the PSWITCH, the problem was, with the ramp power-on the capacitor was charging before the DCDC_IN signal comes to a stable HIGH and by the time it gets there the 1 ms delay in no longer correct (in this case it is less), we have an external reset monitoring circuit and we have connected the PSWITCH to the REST so that in case of performing a reset the capacitor would discharge, we have also increased the tau(RC) value as a back-up solution to make the charging go slower and therefore even if the power-on signal is not sharp enough the capacitor would not be charged to the fullest before the DCDC_IN signal stabilizes at HIGH. We have tested both solutions and they work perfectly.
Best regards.
Hi @Lukaz
I think that the long oscillator startup time issue is covered in this application note: Crystal Oscillator
Troubleshooting Guide, insterestingly, as you described!
Also, we have recommendations for the oscillator in our MIMXRT105060HDUG. See the below snapshot.
It may be only that you add the resistor for startup
Let me know if you are still having issues. Also, take a look at more HW recommendations from our design guide.
All the best,
Diego
