Prototype boards are failing with issue similar to one previously reported (2nd generation same issue)

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
Anonymous
Not applicable

Per the original dialog here: Prototype boards are failing with issue similar to one previously reported (con't)

We still have this issue and it's apparently SDK related.

SDK 2.2.3 on Windows caused very strange errors to occur.  LEDs blinking erratically, resets, etc.  We think there~s something wrong with the fine timer in this SDK version for Windows.

We finally went back to SDK 2.2.2 on Mac.  Now the downloaded code works fine, but the initial problem of about 10% of the boards STOPPING ADVERTISEMENT is back.

How can we solve this problem?  It's always the same boards that fail; it's not random.  For example, if you have 70 boards and about 7 fail, then if you reset/restart all 70, the same 7 will fail again.

This is not a manufacturing issue, because we have used 4 different manufacturers, and different versions of the PCB layout (which follows your guidelines) and always get the same identical problem.

We need to solve this problem badly!!!!

Thanks,

Gil

0 Likes
1 Solution
Anonymous
Not applicable

1)  I don't know why this happens with Windows 10 and SDK 2.2.3 and I wish I knew.  The same code build in the 2 environments behave completely different.

2)  We didn't have a chance to do that, because ......

...... we decided to triple-check on the SDK version that we have installed, to make sure we really have 2.2.2 which supposedlt fixes the "stops advertising" issue:

A) Our installation folder definitely said WICED-Smart-SDK-2.2.2

B) Opening the IDE and clicking on Help > Check For Updates, resulted in "No Updates were found".

C) But clicking on WICED Smart IDE > About WICED Smart IDE, resulted in version 2.1.1 !!!! Huuummmm!!!!

So we decided to uninstall the SDK, but most importantly AND remove all WICED folder from the Mac.  Then we went to the website where the SDK is downloaded and downloaded the SDK 2.2.2 installer gain.  We re-installed it. 

D) Doing C) above again now shows version 2.2.1 (BETTER!!!! but still not right???)

Finally, we reprogrammed 50 failing boards with the same version of code.  These are boards that always failed, even when reprogrammed.  And guess what?   NONE OF THEM HAVE FAILED SINCE!!!!

This makes us very happy, but it uncovers some very serious problems with the Broadcom/Cypress products.  It is ABSOLUTELY unacceptable that the what is seen in A), B), C) and D) above happens.  How are we supposed to keep track of versions if these don't show the correct info?  They need to be fixed!!!  Installation folders need to reflect the correct SDK versions, Get Updates should download updates, etc, etc.

The support team is GREAT though.  You guys never gave up and ALWAYS tried to solve the problem with us.

Thanks and take care!

Gil

View solution in original post

8 Replies
BoonT_56
Employee
Employee
500 likes received 250 likes received 100 likes received

I cannot locate the last posting of your issue and therefore I need you to elaborate a little.

1) Are you using the 20737S module?

2) Did you have a chance to verify your application code on the tag3/4 eval board?

3) Did you apply the "Crystal Warmup" fix? 5000 to 7000 (some suggested)

4) Can you not recover those failed parts?

5) How do you download your code onto the module (I assumed)? cable tools?

SDK 2.2.3 has been around since June 16 and we haven't seen any reported major issue. I installed on a

win7 32-bit machine and this is also my recommendation. Are you using win7?

It is peculiar that the same part will failed after recovering. Did you ensure a good recovery procedure? There

should be a "download.log" file in your build. What does it say on those parts that failed?

MichaelF_56
Moderator
Moderator
Moderator
250 sign-ins 25 comments on blog 10 comments on blog

Adding lsri​ as he is going to try to recreate the issue you are seeing on SDK 2.2.3

Assuming you are using custom firmware and a custom board, would it be possible to recreate the issue on SDK 2.2.3 using the TAG3 development board and one of the sample apps from the SDK?

Also, can you any provide traces/logs?

0 Likes
Anonymous
Not applicable

Hi mifo,

Thanks for the help.

We've tried your sample apps (Simple Advertising, Hello Sensor, Hello Client, etc).  The same thing happens. (Note that some of the example apps actually stop advertising on purpose after a timeout, which is ok).

The problem with trying to recreate in just one board, is that if we take just 1 of our own hardware it may work.  The "stops advertising/scanning" problem only happens in about 1 out of 10 boards, so when we try on about 20 of our boards we find 1 or 2 that stop advertising/scanning after several minutes or several hours, and sometimes it takes 1 or 2 days.  You need a large number of boards.  So, answering your questions, we have not confirmed this on TAG 3/4 eval boards because we don't have 20 or 30 of them.  If you are willing to send us about 50 of the tag 3/4 eval boards, we are willing to run the test.  We can return them to you after the test.

0 Likes

Two things intrigued me here:

1) Why is it that your code get spooky when you migrate to Win? We have similar cases whereby the issues were gone after

migrating to Win with 2.2.3 from Mac and/or from earlier SDK. Can you state your precise Windows environment?

2) For my qn 4 and based on your reply, I realized that you performed a power on/off. What I was referring to was a total

module recovery (as if the code was corrupted). Now, can you collect a few of these "failed" modules and do a full recovery? Let

us know your observation as well as the "download.log" file.

0 Likes
Anonymous
Not applicable

Hi btta,

1) The precise windows we use is Windows 10 build. Is we have seen the Win 10 build behaviour to change for GPIO's.

2) We will check the recovery build logs as well. We haven't done this yet!

0 Likes
Anonymous
Not applicable

Ok.  By "total module recovery", do you mean to reprogram the units?  We will reprogram the failing units and save the download.log file.

0 Likes
Anonymous
Not applicable

1)  I don't know why this happens with Windows 10 and SDK 2.2.3 and I wish I knew.  The same code build in the 2 environments behave completely different.

2)  We didn't have a chance to do that, because ......

...... we decided to triple-check on the SDK version that we have installed, to make sure we really have 2.2.2 which supposedlt fixes the "stops advertising" issue:

A) Our installation folder definitely said WICED-Smart-SDK-2.2.2

B) Opening the IDE and clicking on Help > Check For Updates, resulted in "No Updates were found".

C) But clicking on WICED Smart IDE > About WICED Smart IDE, resulted in version 2.1.1 !!!! Huuummmm!!!!

So we decided to uninstall the SDK, but most importantly AND remove all WICED folder from the Mac.  Then we went to the website where the SDK is downloaded and downloaded the SDK 2.2.2 installer gain.  We re-installed it. 

D) Doing C) above again now shows version 2.2.1 (BETTER!!!! but still not right???)

Finally, we reprogrammed 50 failing boards with the same version of code.  These are boards that always failed, even when reprogrammed.  And guess what?   NONE OF THEM HAVE FAILED SINCE!!!!

This makes us very happy, but it uncovers some very serious problems with the Broadcom/Cypress products.  It is ABSOLUTELY unacceptable that the what is seen in A), B), C) and D) above happens.  How are we supposed to keep track of versions if these don't show the correct info?  They need to be fixed!!!  Installation folders need to reflect the correct SDK versions, Get Updates should download updates, etc, etc.

The support team is GREAT though.  You guys never gave up and ALWAYS tried to solve the problem with us.

Thanks and take care!

Gil

Anonymous
Not applicable

Hi btta,

Thanks for the response:

1) Yes.

2) The problem is that if we take just 1 of our own hardware it may work.  The "stops advertising/scanning" problem only happens in about 1 out of 10 boards, so when we try on about 20 of our boards we find 1 or 2 that stop advertising/scanning after several minutes or several hours, and sometimes it takes 1 or 2 days.  So, answering your questions, we have not confirmed this on TAG 3/4 eval boards because we don't have 20 or 30 of them.  If you are willing to send us about 50 of the tag 3/4 eval boards, we are willing to run the test.  We can return them to you after the test.

3) Yes.

4) After the failed boards "stop advertising/scanning", if we reset them they will restart code execution and will advertise again.  But they will eventually stop adverting again in minutes, hours or a couple of days.

5) We compile and download on WICED with Mac, SDK 2.2.2 and USB-to-UART (FTDI) cable.

We use Mac.  When is SDK 2.2.3 going to be available for Mac?  This is important for us, because we tried rebuilding and downloading the code with SDK 2.2.3 on Windows and saw very strange behavior of the code (fine timer was not working properly).

Yes, it's very interesting that it's always the same boards that keep failing.  We have tried 4 different manufacturers (because we initially thought it as an assembly issue) and the failures occur with all of them.

Now that you asked these questions, I'm wondering one thing.  Is the problem caused by an unseen/unreported programming issue somehow?  The programming always goes to the end, successfully; but was it really successful?  So I'm wondering that maybe we should take the failing boards and reprogram them, then run the test again.  We haven't done that.  Huuummm....

sambhav

0 Likes