K10 Slow Power Application of VDD causes lockup

[howard_lehman]

Experiencing power lockup of K10 LQ12 part.   Over heats immensely. 

Trying to make it clear to community, this is mature product in Alpha Beta production.  The strange problem surfaced after a PCB layout change to other areas of the PCB, not directly to I/O of MPU.    Prior to this revision, there has been no occurence of this problem. 

The term "lockup" means that upon power application to the device it does not start running code, instead it gets too hot to touch.  However, once removed from the machine, which difficult an compromising to the quality of the assembly, the MPU powers up and runs code properly with no obvious issues.  This is not a solder short.  

The NXP Service Support suggested, rightly so how to isolate the problem.   However,  the problem occurring in 3 different machines that have other ongoing design verification work,  can not become test beds for what if approach to this problem. and hence is impractical. ways to isolate condition which were impractical.

[austinhere]

A slow rising power supply may cause the MCU can not reset properly. You may need to add an Reset IC to let the MCU leave reset state after supply voltage reached working area.

[howard_lehman]

Hi Austin here,

Yes we are using MCP1315-29 WDT. It has worked great! Feeding /Reset pin through 1k ohm.

Reset occurs approx.. 200ms after power within limits.

[kerryzhou]

Hi Howard Lehman，

  What's the detail mean of your:power lockup of K10 LQ12 part

  Do you mean, the K10 chip is locked, and can't find the ARM core?

  Please check your hardware, whether the solder have the short problems, please also check your power supply, whether the 3.3V is Ok or not?

  Your problem should still in the hardware side.

  Any updated information, please kindly let me know.

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

[howard_lehman]

Hello Kerry,

Thanks for responding to an open issue and call for support.

I hope you can appreciate my situation and that it is very difficult to explain a fundamental issue in all the logistics which is now the new standard in tech support. I will bet you have never seen or heard of this behavior.

How can we effectively communicate?

Regards,

Howard Lehman, EE- Contractor (extension: 27804)

Hologic Inc,

36 Apple Ridge Rd.

Danbury, CT

203.702.7804

[kerryzhou]

Hi Howard Lehman,

  You mentioned Service Support, do you already create the ticket in the www.nxp.com, and get some support?

  Talkback to your K10 technical questions, do you mean your chip is locked and can't connect?

  If yes, please check this document at first:

https://community.nxp.com/servlet/JiveServlet/download/857127-1-396709/Kinetis%20Lock%20issue%20anal... 

  I think you can test your K10 smallest system hardware, program function.

  If you meet the technical issues, please keep patient, I totally understand you, let's analyze it step by step.

 1. Check smallest system, whether the design have problems or not.

 2. Check hardware solder, make sure the solder is good

 3. check the power supply, make sure the voltage is correct.

 4. use JLINK connect your chip through the SWD, check whether the ARM core can be found or not.

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

[howard_lehman]

Hi Kerry,

My case was just closed. I don’t think they understand all the details and help I needed.

Upon reviewing your “Blank” link.. I don’t have the opportunity to see RESET and Power behavior in system thus far, due to demands by others on the prototype complete systems. There are very expensive and probed testing for an occasional event are not high priority.

In “Blank” there is no mention of extreme device heating or the “K10FX part we have and VDD slew rate, as being a concern. I believe the internal power monitors are at fault and core starts running doing crazy port writes that end up in over clocking internal hardware and causing heat ,, 80°C heat at ambient board of 25°C

Regards,

Howard Lehman, EE- Contractor (extension: 27804)

Hologic Inc,

36 Apple Ridge Rd.

Danbury, CT

203.702.7804

[kerryzhou]

Hi Howard Lehman,

  Please give me your case number, I need to check more details about your question, because until now, I am also not very clear about your detail questions.

   You said your K10 LQ12, power lockup, Over heats.

    How many chip on your side meet this problem now?

   Do you refer to the hardware design to design your smallest system？

https://www.nxp.com/webapp/Download?colCode=KQRUG 

  Do you add the 4.7k to 10K pull up in the reset_b pin, and 0.1uf capacitor to ground?

  What's your core clock which you have configured?

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

[howard_lehman]

Kerry,

Glad to here again. I have removed some EMAIL BLOCKING for your inbounds.

I’m dealing with a family of 10 parts installed on assemblies, at least 3 have exhibited the condition. I can not easily swap parts and try again, this is impractical at this point of the design. I’m trying to improve hardware design to minimize potential influencers. However, this is all guess work thus far. The design has proven stability over a long period, but a recent board revision to what I have shared already has brought on this problem.

Case Number : 00262135

[kerryzhou]

Hi Howard Lehman,

 1. About your case 00262135

   After checking it, I find our two colleagues already give you a lot of suggestions, do you try it on your side?

  I find my colleague reply you on 3/6/2020, but it seems you didn't give any feedback, they still waiting for your reply, could you tell me why you don't follow it any more?

2. Talkback to your question

   You mentioned: I’m dealing with a family of 10 parts installed on assemblies,  at least 3 have exhibited the condition.

    Do you mean, other 7 boards are working perfectly, with the same code, the same board design, the same K10 chip, right? Just 3 boards meet the overheat problems?

    Could you please tell me your K10 LQ12 detail part number, and your maskset?

    With the 3 issue boards, do you try the case engineer suggested way: 

load a simple project that only set one gpio,with this you can check if the device is getting hot or not, if not then you can add
other module like uart, and continue the test Also can you check if there is not any manufacturing issue,like a solder short, missing part…
If you can not perform a swap test, then I believe the best option to find the problem is testing module by module

      Any updated test result?

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

[howard_lehman]

Kerry,

I am not able to have any software/programmers work on test code. That development cycle has past over one year ago. The software available is only Normal and Bootloader, which still enables CAN communications. Unfortunately, to load this code, the PCB has to removed from the machine, which takes about ½ hour, and add’s risk to the machines harness and other functioning parts. Since the machine must function after the time to remove the PCB, only a known good PCB can be replaced and will not have trouble, with the same code. Please consider that this software staff is doing higher functioning tasks and this part is rarely code revised. Generally minor errors in function, logic reversal, or CAN function changes.

Sorry I have limited time here…

1. No reply, was probably from email company screening them, so I don’t see them for days.

2. Mask? I have no idea what this is. I gave the full part number in the case but have not supplied date code. I can do this for you if this is what you need. If these are the numbers after a 4 digit “5N96B” number there is no obvious pattern.

XNCTAZ1831J, XNCTAH1806A, XNCTAH1806D, XNCTAH1806C, REPRESENTS ALL PARTS I IN MY IMMEDIATE POOL. --- 1ST ONE IS A HOT ONE, THE OTHER IS IN A MACHINE, AND WILL TAKE TIME TO SEE THE NUMBERS.

[kerryzhou]

Hi Howard Lehman，

   If your board already assembled and you can't test the software and the hardware details, it is really very difficult to find the root problems. You said, the same code, the same PCB, but just in 10 product, 3 board have problems.

   This development cycle is past one year, when you design this project, how many boards you have assembled? Any related problems? I think it's better remove the issue board, and replace it with the working PCB, then you can let your related software and hardware team to check more details about the hot problems.

   About the Maskset, your 5N96B is the maskset number, I checked the errata:

 https://www.nxp.com/docs/en/errata/KINETIS_K_5N96B.pdf 

  I didn't find any related issues.

  And your side already has 7 boards working in 10 boards, to be honest, I also suspect your 3 issue board hardware factors, eg, the solder short.

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

[howard_lehman]

Hi Kerry,

Thanks for reply.

This is Not a solder short. I’m on the hardware team, and work with software team.

The changed aspects of the MPU hardware the biggest potential causes. It is remotely possible that this is board fabrication problem, but not a hard failure, one that can be reproduced.

I will now share more details, as your line of suggestions is not addressing my concerns. What if the power application of Vdd & VBAT looks like attached.

[kerryzhou]

Hi Howard Lehman,

   Thank you for your updated information.

   Do you mean, when you meet this Vdd & VBAT wave, then your K10 is over heats immensely? What about the code function, whether it works or not? I just want to check whether chip is boot successfully or not.

   BTW, could you please also add the reset pin wave in your vdd&vbat wave? And provide the working board related vdd&vbat wave?

Kerry

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------