BLE transparent mode firmware for STM32WB55VGQ package.

[rapoorv]

Hello,

Using custom firmware on an Inhouse designed board, the BLE output seems unmodulated as shown in the picture attached. But using the Transparent mode firmware on a Nucleo-Board (controlled via the Cube monitorRF) the modulation seems good (also attached)

So,

1. Is there any option to tune the modulation further other than default ?

2. Can someone provide the Transparent mode firmware for the STM32WB55VGQ package ? This can help in FCC testing.

This picture shows the sidebands when tx is ON on the 1st channel.

This picture shows there are no sidebands on the Nucleo Board which are coming in the above pic.

[Lubos KOUDELKA]

Hello,

Using SW setting you can tune HSE frequency and TX output power, but then it's matter of filtering, antenna and antenna matching on your custom board.

You can use BLE_TransparentMode project from STM32Cube_FW_WB repository intended for P-NUCLEO-WB55.Nucleo on any STM32WB55 package. You just need to reflect which interface (UART, LPUART, USB) and I/Os are used for communication with PC on your custom board and change it accordingly in your BLE_TransparentMode project.

Best regards,
Lubos

[rapoorv]

Hi,

The porting of the code was tried but we cannot control it via the CUBEMonitorRF app. 
We used the same code and generated the TX signal manually at a fixed frequency of 2402MHz, but we could still see the sidelobes in the spectrum analyzer whereas it does not appear on the Nucleo-Dev kit.

Can someone in ST do the porting of this code onto the package for STM32WB55VGQ and test it with the MonitorRF app. 

Thanks

[Lubos KOUDELKA]

Hi,
as I read your your feedback, porting of your code probably was not complete, otherwise you should be able to control your app using STM32CubeMonitorRF. Only think you need to port is to set (LP)UART and right pins (LP)UART is using on your board. Code is not package specific, but there is lot of combinations of UARTs and pins you may use on your custom HW, we're not able to cover them all in examples. What pins are used on your side? I can then guide you to the modification you need to do.

But the root cause here is different - as you were able to see using the fixed frequency signal you generated - you need to fit your RF filtering and matching to your HW board, no SW modification will fix it.

[rapoorv]

Well if you see in that way, we are testing at the RF pin of the controller with no further connections. If at that point we observe the side lobes, impedance matching circuit is not gonna help neither a filter. 

Now if we use a Nucleo Dev board (MB1355C), we do not see any sidelobes if we probe at the RF pin of the controller.

As for the Transparent mode firmware porting, would be great if you could provide some guidance .. 

[Lubos KOUDELKA]

What about your HSE and input power stability? Have you verified it?
For test would be good to power your MCU from stable source with proper decoupling, or some battery. Regarding HSE, are you able to switch for stable signal generator for test?

[rapoorv]

Checked the HSE stability and 3.3V line (near to RF as well at VDDmain). All good there.

Checked it from an even better source with lesser ripples of 20mV on 3.3V line. The problem persists.

Whats the 2nd statement about the signal generator?

[Lubos KOUDELKA]

2nd statement about the signal generator - RF use HSE 32 MHz as source for carrier - HSE need to be precise and stable, HSE HW design is sensitive part on STM32WB. Using signal generator with stable 32 MHz signal to bridge HSE crystal may help to verify if the issue was HSE. You can verify HSE also using other methods, for example measurement on MCO using precise clock counter.

[rapoorv]

HSE frequency stability at room temperature is tested for accuracy. Its a +10ppm part. Have tried to tune in the HSEGMC down and up. It only shifts the center frequency and does not helps with the Side lobes coming.

[teja_111]

Hello @Lubos KOUDELKA @rapoorv 

Did you managed to bring up the BLE transparent mode firmware for STM32WB55VGQ6 package?

We are also trying to port the nucleo board example of the transparent mode to VGQ package, with LPUART pins PB5 - TX and PA12 - RX. But have no success and Hangs in the Hard Fault handler.

Is there any memory constraints that we need to change in the generated Flash file for the transparent mode to work?

This would be very helpful for testing with RF tool. Please provide some guidance on this topic.

Best Regards

Teja