• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 79 subscribers
  • Views 1327 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Flexing PCB with VCORE set too high causes PUC (0x20 PMM password violation)

Nason Tackett
Prodigy 70 points
Other Parts Discussed in Thread: MSP430F5515

After about pulling my hair out and suspecting that there was a manufacturing problem (broken vias, etc) with the PCB, I scoped the power pins, and disabled all I/O and still had the issue of the MSP430F5515 rebooting each time pressure was put on the PCB (even after reflowing the solder and also on multiple boards)...

Each time the chip rebooted, the power on all the power pins was rock steady, but 0x20 (PMM password violation) was in the SYSRSTIV. After doing a little digging, I realized that I had forgot to drop my core voltage before switching to 24 Mhz MCLK on the DCO... I didnt even suspect this would be the cause of the rebooting, but knew that I should have been dropping the voltage anyway, and after I made this change, flexing the PCB (no matter how hard I try) does not cause the reboot. Before I made this change, I could make it reboot on demand anytime and on multiple PCBs. It seemed to be the most sensitive next to the capacitor for the VCORE pin. The only thing I could figure is flexing was causing some kind of change in capacitance and pushing the VCORE just over the limit to cause issues? Any ideas? 

I post this because I have a hard time understanding why having the VCORE too high could cause such an issue with flexing the PCB... (the application this will be used in will have a lot of mechanical vibration so this would have been a serious issue...)

  • 0
    Guru 227245 points

    Sorry, I don't understand what you're saying. You drop (lower) your core voltage before switching to highest MCLK frequency?

    AFAIK (and as far as the datasheets and hte users guides and also common CMOS experience tell) you'll need the highest core voltage for highest speed. 24MHz requires PMMCOREV = 3.
    So I wonder how your schematics is. Do you have the required 470nF ceramic cap on VCORE? Did you connect VCORE to anything else (which you shouldn't)?

  • 0 in reply to Jens-Michael Gross
    Prodigy 70 points

    I apologize, I should have said I'm increasing the voltage...  I am changing PMMCOREV = 3, so I am increasing the voltage before I switch MCLK to 24Mhz. I do have the 470nF cap and it is very close to the VCORE pin and nothing else is powered from this. 

    One possibility is that the 470nF cap is faulty (possibly cracking and becoming intermittent). I say this because I have come across some PCBs that do not have the issue when flexed. But the ones that do have the issue seem to be fixed after I change PMMCOREV = 3. So maybe trying to run at 24 Mhz with PMMCOREV = 0 and a flaky 470nF cap could explain this? I will do further testing to see if my theory on the caps is correct... These boards are potted otherwise I would have tested this already.

  • 0 in reply to Nason Tackett
    TI__Guru 70346 points

    Even if your cap is perfectly fine - running the device at 24MHz with PMMCOREV = 0 is WAY out of spec, so it does not surprise me that you'd see resets. This is a violation of voltage vs. MCLK freq - as you can see in the datasheet on p. 52 Figure 1, the highest allowed MCLK with PMMCOREV = 0 is 8MHz, and you are running at three times that. A common issue with voltage vs. MCLK spec violations is that the device will act erratically or reset, so I think your behavior is expected. As long as you raise your core voltage to PMMCOREV = 3 before making MCLK 24MHz, you should be fine because you'll be within the datasheet spec.

    Regards,

    Katie

  • 0 in reply to Katie Pier
    Prodigy 70 points

    Thank you Katie. I agree that erratic behavior is expected with that voltage and frequency. I just wanted verification on the behavior that I was seeing. Thanks for the info!

**Attention** This is a public forum