Skip to main content
Associate III
Nov 21, 2026 9:23 PM UTC
Solved

STM32WB1x BLE Full Extended Stack V1.25.0 – USART DMA streaming issue

Hello ST team,

I am using the B-WB1M-WPAN1 (STM32WB1MMC) board with the BLE Full Extended Stack.

My application is based on BLE_DataThroughput and operates as a bridge between an STM32G473 and a smartphone.

Data can be transferred from the STM32G473 to the STM32WB1MMC using either USART1 DMA or SPI DMA, and then forwarded through BLE notifications (244 data bytes + CRC8).

The application works correctly with STM32Cube_FW_WB V1.24.0 and its BLE stack.

After updating the STM32WB1x BLE Full Extended Stack to the version supplied with STM32Cube_FW_WB V1.25.0, I observed the following:

  • Both SPI transmission modes continue to work.

  • USART1 DMA streaming no longer works correctly.

  • BLE advertising, connection, initial handshake and commands from the smartphone still work.

  • USART1, DMA1 and DMAMUX registers appear correctly configured.

  • Increasing CFG_TLBLE_EVT_QUEUE_LENGTH from 5 to 10 did not resolve the issue.

I also tested the original working V1.24.0 application without modifying its source code, using the new V1.25.0 wireless stack. The result was the same: SPI worked, but USART streaming did not.

My question:

Are there any changes in the STM32WB1x BLE Full Extended Stack between V1.24.0 and V1.25.0 that require modifications to CPU1 application initialization, BLE event handling, transport-layer buffers, or notification processing?

I am particularly interested in changes specific to the STM32WB1x binary stack, rather than STM32WB55.

Thank you for your assistance.

Associate III
Nov 22, 2026 6:15 PM UTC

SOLVED – My apologies to the ST team!

Dear ST team,

I am happy to report that STM32WB1x BLE Full Extended Stack V1.25.0 works correctly in all four modes of my application.

The problem was entirely on my side.

When switching between SPI and USART, I changed the conditional compilation settings in both microcontrollers but forgot to reposition the physical jumpers on my adapter board.

As a result, the USART communication lines were physically disconnected.

After correcting the jumper positions, everything worked immediately.

I have now tested all four operating modes with STM32Cube_FW_WB V1.25.0:

Operating mode Average throughput
SPI DMA 89.40 kB/s
SPI DMA + hardware CRC-8 + ACK/NACK 94.64 kB/s
USART DMA 85.34 kB/s
USART IDLE 81.27 kB/s

All measurements were performed using ST BLE Toolbox, with 245-byte BLE notifications and 2M PHY.

I have attached four screenshots showing the results.

My sincere apologies for raising a false alarm about the V1.25.0 BLE stack.

I must now sprinkle ashes on my grey head! :)

Thank you to the STM32WB development team. You have done an excellent job, and I am very pleased with the performance of this device.

Since the reported problem was caused by my hardware configuration rather than an ST software issue, I would appreciate it if a moderator could close or remove this topic, whichever is more appropriate.

Best regards.

1 reply

pm_interrupt_vectAuthorBest answer
Associate III
Nov 22, 2026 6:15 PM UTC

SOLVED – My apologies to the ST team!

Dear ST team,

I am happy to report that STM32WB1x BLE Full Extended Stack V1.25.0 works correctly in all four modes of my application.

The problem was entirely on my side.

When switching between SPI and USART, I changed the conditional compilation settings in both microcontrollers but forgot to reposition the physical jumpers on my adapter board.

As a result, the USART communication lines were physically disconnected.

After correcting the jumper positions, everything worked immediately.

I have now tested all four operating modes with STM32Cube_FW_WB V1.25.0:

Operating mode Average throughput
SPI DMA 89.40 kB/s
SPI DMA + hardware CRC-8 + ACK/NACK 94.64 kB/s
USART DMA 85.34 kB/s
USART IDLE 81.27 kB/s

All measurements were performed using ST BLE Toolbox, with 245-byte BLE notifications and 2M PHY.

I have attached four screenshots showing the results.

My sincere apologies for raising a false alarm about the V1.25.0 BLE stack.

I must now sprinkle ashes on my grey head! :)

Thank you to the STM32WB development team. You have done an excellent job, and I am very pleased with the performance of this device.

Since the reported problem was caused by my hardware configuration rather than an ST software issue, I would appreciate it if a moderator could close or remove this topic, whichever is more appropriate.

Best regards.