- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi, we are using CYW20706+Wiced6.4 for an important project, after days of pressure test, we found a tricky problem:
In a fast booting senario (power on/off hundreds times ), part of SS section (2~3KBytes from address 0x0000) might be written to 0 , so the chip will boot to HCI mode as it think the SPI flash is invalid, although the application area (start from 0x4000) is still there.
This is a huge risk for us as when SS corrupted , the chip can not work at all !
Can anyone help me ?
HELP!
THANKS!
Solved! Go to Solution.
- Labels:
-
WICED Studio Bluetooth
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
For the abnomal chip, i notice that it boot to HCI mode (send hci reset command 01 03 0C 00 to chip , chip response 04 0e 01 03 0c 00 ), so i remove the SPI flash and attach to an exteral MCU (i.e. STM32F4) for content reading, and it turn out that the SS flash area corrupted.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @towa_2281366 ,
Static Section of the external flash is used internally by the chip FW. How do you notice that 0 is getting written to SS? Are you seeing any change in application behavior afterwards?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
For the abnomal chip, i notice that it boot to HCI mode (send hci reset command 01 03 0C 00 to chip , chip response 04 0e 01 03 0c 00 ), so i remove the SPI flash and attach to an exteral MCU (i.e. STM32F4) for content reading, and it turn out that the SS flash area corrupted.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Add some more details:
1. In nomal circumstance, the chip works fine, i.e. Using an external MCU's gpio to control a LDO for powering on/off 20706 periodic, after test 100000 times , chip boot normal.
2. Disaster happens in a "fast booting" senario, powe jitter may exist in the end product when booting, i doubt that the boot loader of chip may have write actions to SS area.