STM32WB15 Dissapointment

[PLee.3]

Hi,

I just have to voice my opinion on STM32's BLE STM32WB15 solution I have got, out of the box (Nucleo-WB15CC) I have spent many hours now just trying to get a simple example started just advertising to a mobile but I still can't get it too detect. Apparently reading around I have to flash "Binary" files straight to the MCU before I can get an app to work, this I just think is too much for a user (Adding that I can't seem to install STM32CubeProg). Compared to an ESP32 on espressif framework where you can get BLE up and running in a matter of minutes. I am now stuck at a stage where I have to rely on the ESP32 framework to provide a solution for my company as the STM32 equivalent is far to complexed. Add to that if this was an industrial type solution we would have to flash each MCU twice just to get them up and running! 

Does anyone else feel this way?

[AScha.3]

Hi,

just from my limited experience, i would say:

+: on Arduino or ESP framework you get complex things done with just some calls. 

-: you dont know, whats really going on, in detail. (No documentation, whats chip/register settings are.)

+: on STM32, you have good documentation ( some thousand pages...) and you know, what you do.

-: until you get something complex working, it might need "learning" many hours (200+...)

So the real question is: can you use the ESP thing with the calls you need and dont need more detail - ESP !

or need some precision action, in detail you have to know, what happens - STM !

And ESP is (maybe) world leader in simple IOT , WiFi or BT access to ... , 

while STM is (maybe also) world leader in industrial, embedded controllers .

So for a good PWM motor controller, i would choose STM.

To connect wireless something, i would choose ESP.

Or , as i do, both : STM is the master for hardware , communicating with the ESP for the network things.

[Andrew Neil]

---

@AScha.3 wrote:

-: until you get something complex working, it might need "learning" many hours (200+...)

---

Exactly this.

Bluetooth is complex - it does have a steep learning curve.

[Andrew Neil]

---

@PLee.3 wrote:

I am now stuck at a stage where I have to rely on the ESP32 framework to provide a solution for my company 

---

Why is that a problem for you?

For a commercial product, the cost of development - including the time for learning - has to be considered.

So, if an ESP (or whatever) does the job, then use that!

---

@PLee.3 wrote:

Add to that if this was an industrial type solution we would have to flash each MCU twice just to get them up and running! 

---

Not sure what you mean by that?

In an "industrial type solution", you'd have a programming system that would take the two files and program them both as a single operation.

[PLee.3]

Still almost 2 months in and can't get my BLE STM32 device to properly advertise some custom characteristics... The 2x flashing is the binary flash (plus BLE Stack Enable) + STM32 Program. There either needs to be some better literature/examples or I will just carry on using ESP32 or Nordic device...

[Tesla DeLorean]

Problem is, they are a Multi-$B chip company, looking to do business with other $B companies with a sufficiently wide and deep enough bench of engineers with domain knowledge to get their own job done. They are not about "user" experience, that's the job of the "developers" using the hardware and writing the software for their application / use-case.

May be you're doing enough business that you can get an FAE to work with, or that they can suggest specific contractors. Start at a business level working with your local sales office or distributor.

Paranoia with security also makes the updating processes non-trivial, by necessity. Perhaps you can leverage FUOTA as a means to deliver incremental code changes, or make a loader/monitor capable of taking RAM base applications or applets whilst you develop and debug your solution.