STM32F4RGT6 random initialization issue

[zexx86]

Posted on June 21, 2015 at 14:51

Hello again!

We have designed the board with this processor, it works perfectly.

But in 1 : 100 times it seems as there is some weird condition and the processor wont initialize at all.

If this situation happen, then I can re-power up the board as many times I want and it never boot up.

Only when I wait few minutes it works again as without any issue.

Then it works very well for next XXX power ups.

I have found that when I discharge the capacitors then it works immediately again.

So definitively there is some issue with the capacitors.

I suspect that 47uF tantalum capacitor is not suited correctly.

Please check my scheme.

I have thought there is issue with BOOT0 pin, but it is not the case.

Can you see something wrong in my design?

It is not a problem with one particular board as the issue is happening with all assembled PCBs. So there is definitively design issue.

Thank you very much for your great support!

[Tesla DeLorean]

Posted on June 21, 2015 at 15:24

Tying NRST to 3.3V is a very bad plan. The processor tries to drive this low, it's a bi-directional pin.

Consider also the initial state of BOOT1

[zexx86]

Posted on June 21, 2015 at 16:43

Oh, thank you very much. I didn't know that.

I have disconnected the NRST pin - we will see how it will behave. I have reviewed datasheet and I now understand why it should not be connected to VDD.

Also it is recommended to use capacitor to avoid parasitic resets.