Hi TI employees(s),
We’ve developed first article BLE radios using the CC2541 module. Our early firmware release was developed using the Blue-Giga BGscript, but going forward we want to switch to a full ‘C’ development environment. We purchased the IAR 8051 IDE and are using the TI BLE-Stack 1.4.0.
In brief, our product scans on all three advertisement channels and then sends the resulting MAC addresses and RSS across the UART at 115k baud to our server.
We have encountered some issues with the TI BLE stack.
Specifically, we’ve had issues receiving data across the UART while scanning. More importantly, starting a scan using the TI BLE stack, calling GAP_DeviceDiscoveryRequest() with DEVDISC_MODE_ALL, activeScan=FALSE, whitelist=FALSE, GAP_GEN_DISC_SCAN_INT=160, and TGAP_GEN_DISC_SCAN_WIND=160, it returns less samples and devices than the Bluegiga stack. Typically in 60 seconds of scanning, 243 samples and 14 unique devices are returned, compared to 416 and 21 with the Bluegiga stack.
We noticed a difference in the receiver sensitivity of the two stacks. The Bluegiga stack commonly returns RSS samples as low as -94 dB, while the lowest RSS discovered by the TI BLE stack seems to be -91 dB (I have set the TGAP_SCAN_RSP_RSSI_MIN=-127).
Also, because the TI stack doesn't return devices multiple times during a scan unless the advertisement data changes, we are forced to restart the scan more often than we would like.
We are not sure if any of these issues are show stoppers, but we certainly want to understand TI’s commitment to this product before moving forward.
When is the next version of this product slated for release?
Are any of the issues that we are encountering being addressed?
Is there something in the works that we could beta test?
Is it possible to get a custom release of the software stack?
Thanks,
Dalton
