Why SPI in BlueNRG hungs suddenly after hours of correct work?

Piotr Romaniuk · ‎2018-05-11

Posted on May 11, 2018 at 12:33

Hi,

In my device I use BlueNRG-2 working as Network Processor (DTM firmware is loaded there).

It cooperates with STM32 where is my application. Processors communicates each other via SPI interface.

It works fine for long time but suddenly BlueNRG-2 stops responding.

STM32 change BLE_CS pin to active state (low) but BlueNRG never rise BLE_IRQ.

It may happen after half hour or couple of hours.

The application works periodic in two phases:

a) dicovery phase,

b) connection phase.

In discovery phase the device scans available external LE devices.

Then it connects to one of them and reads some characteristics value.

After that it disconnects and returns to scanning phase, and starts from the beggining.

It looks like a bug in BlueNRG firmware (provided by ST Microelectronics in Dev Kit).

What can be a reason of such problem?

How to debug it and get rid of it?

King regards,

Piotr

#bluenrg-2 #spi #hung

Piotr Romaniuk · ‎2018-05-30

Posted on May 30, 2018 at 08:36

Observation of communication between BlueNRG GUI SW and BlueNRG-2 is nor required on this stage of problem analysis.

I already checked that:

1) chip goes to sleep

2) waking up signal and configruation is correct (it works for most of the time)

3) problem exists only for specific delay 30us. It works for shorter or longer delays. By the way, 30us is close to 1 cycle of low speed oscillator.

It looks like hazard issue due to missing or incorrect synchronization inside the chip.

From the application perspective problem is critical, communication with the chip is lost only reset helps.

Kind regards

Piotr Romaniuk

Gerardo GALLUCCI · ‎2018-05-30

Posted on May 30, 2018 at 19:49

Hello Piotr, Michael,

we are sorry for 2 weeks passed, and for issue you found. Please consider that we are investigating the issue, so, we are fighting with you against it. Antonio asked about stack version just to be sure is not something already fixed into older version. If nobody from ST is still proposing a solution, well, this is because unfortunately there is still no news to share with you. I'm sorry again.

So, generally speaking, please give us all required steps to reproduce the issue, and give us some time to fix and/or propose a workaround is/when possible. Every question from ST, and every reply from you, even if it seems trivial, helps a lot.

Best regards

jerry

ps: we are people, and psychological effect usually helps a lot. I hope you can understand. You are not alone!

Michael Krechko · ‎2018-05-31

Posted on May 31, 2018 at 07:38

Hello Gerardo,

what is the weather like in Italy?

To be clear I meaned support requests via web form, that is processed by deep learning AI algorithm, I think. And it generates 'try to use SDK', 'try to read docs', 'try to read community'))

There is absolutely another situation with real ST Engineers, usually they really try to help.

Thank you for your efforts!

Best regards

Michael

Piotr Romaniuk · ‎2018-06-04

Posted on June 04, 2018 at 23:15

Hello Gerardo,

I am very disappointed by ST support and I agree with Michael opinion on this matter.

I opened support request on st.com site on this subject as well. The first suggestion was: of course, the question about DK version and method that I used for measuring 30us delay. After I rejected this as a solution, that obviously was not the one, I have been contacted by email just 'to speed up discussion'. This time questions were closer to my problem, but showed that the person did not read my posts. It really looked like AI response:

'

I you can share sleep instructions used

'

I tried to be polite and explained again, but my answer is left without any further contact from ST support (already 17 days). I sent another email with question if my previous one was received, but this time I asked for confirmation of receiving. No response, like it went to /dev/null. My support request in my profile was closed as well.

I also doubt that you read these posts, because you are asking me for steps to reproduce the problem. I already did it, and explained in details where is the problem and how it is manifested. All you need is ST example on BlueNRG and some application. I just expected more effort and some responsibility from chip manufacturer. Unfortunatelly, it resembles 'AS IT IS' rule well known from software world.

Kind regards,

Piotr Romaniuk

PS

In next design I need to check another chip for BLE, for sure from ST competitor. Maybe it will have better support.