CC2530: At 2v, CC2530 factory resets

Jack Loop

Part Number: CC2530
Other Parts Discussed in Thread: Z-STACK,

Hi,

I've noticed that when I bring my power supply down to 2v, my ZED has trouble communicating with the ZC, and then it strangely drops connection with the ZC and starts doing beacon requests (I assume this is because it factory resets itself). I re-powered the ZED with 3v, and it did not rejoin to the ZC, but continued doing beacon requests. Why does the ZED do a factory reset when the voltage is low?

over 7 years ago

0 YiKai Chen over 7 years ago

Guru 735685 points

When CC2530 does reset and doing some NV operation in low voltage, it might damage NV. As I remember, Z-Stack already implement protection to this. If you still have this trouble, I suggest you to implement some kind of protection from application side.

0 Jack Loop over 7 years ago in reply to YiKai Chen

Guru 17870 points

Where is the code for the Z-Stack's protection implementation?

0 YiKai Chen over 7 years ago in reply to Jack Loop

Guru 735685 points

There is "if ( !OSAL_NV_CHECK_BUS_VOLTAGE )" in osal_nv_item_init() and osal_nv_write().

0 Jack Loop over 7 years ago in reply to YiKai Chen

Guru 17870 points

I see "return NV_OPER_FAILED" happens when the voltage is too low. However, the ZED still gets a factory reset. And I see it is stuck, doing nothing, at a constant 10mA. Why is this so?

0 YiKai Chen over 7 years ago in reply to Jack Loop

Guru 735685 points

I have no idea about why ZED still gets a factory reset. Ryan Brown1 Do you have any idea?

0 Ryan Brown1 over 7 years ago in reply to YiKai Chen

TI__Guru**** 217467 points

Hi Jack,

Are you able to demonstrate this with a TI example? Please provide the steps necessary to replicate the issue. Are you sure that your power supply give a steady 2V and does not go any lower, and does the same occur when powering straight down from 3 V and restarting? From what operation is NV_OPER_FAILED returned?

Regards,
Ryan

0 Jack Loop over 7 years ago in reply to Ryan Brown1

Guru 17870 points

I use GenericApp ZED. My power supply gives steady 2v. When I power from 3v, I don't see this problem. NV_OPER_FAILED is only returned when voltage is below VDD_MIN_NV.

0 Jack Loop over 7 years ago in reply to Jack Loop

Guru 17870 points

Ryan Brown1 Hi. Any update on this?

0 Ryan Brown1 over 7 years ago in reply to Jack Loop

TI__Guru**** 217467 points

I was able to replicate this on my setup. OSAL_NV_CHECK_BUS_VOLTAGE is defined as HalAdcCheckVdd(VDD_MIN_NV) where VDD_MIN_NV is (VDD_2_0+4). VDD_2_0 is 74 and the additional 4 is for a 5% margin over minimum to survive a page erase and compaction. What this means is that 2.1 V is required to pass the OSAL_NV_CHECK_BUS_VOLTAGE. Since you are using a value underneath that voltage then the NV is erased during a write or item initializaiton. You could remove the "+4" from VDD_MIN_NV to resolve the issue but TI recommends using over 2.1 V for the reason provided earlier.

Regards,
Ryan

0 Jack Loop over 7 years ago in reply to Ryan Brown1

Guru 17870 points

Ryan Brown1 I've left VDD_MIN_NV as the default value already (VDD_2_0+4). I'm just not sure why my device would still factory reset if the voltage is below this though. I tried increasing the margin, but I still get this problem. What might be causing this?

0 Ryan Brown1 over 7 years ago in reply to Jack Loop

TI__Guru**** 217467 points

There is a bug where some items in NV memory are deleted and set to their default value if the device is booted or running at a voltage close to VDD_MIN_NV. This is because there was no check done to confirm that the NV Item needed to be initialized to the default value. A fix is being formulated to add an NV check during initialization to set the NV item to the default value only if it was actually initialized and did not exist previously. An update will be provided after more information is available.

Regards,
Ryan

0 Jack Loop over 7 years ago in reply to Ryan Brown1

Guru 17870 points

Ryan Brown1 Hi. Any update on this?

0 Ryan Brown1 over 7 years ago in reply to Jack Loop

TI__Guru**** 217467 points

We are currently testing the workaround and will provide an update when it is validated.

Regards,
Ryan

0 Jack Loop over 7 years ago in reply to Ryan Brown1

Guru 17870 points

Ryan Brown1 Hi. I just wanted to ask if you've been able to validate the workaround as of now.

0 Ryan Brown1 over 7 years ago in reply to Jack Loop

TI__Guru**** 217467 points

Hi Jack,

Apologies for not replying more quickly, a fix has been implemented but as it affects both the Z-Stack core and application you will need to wait for the release of 3.0.2 planned for this week.

Regards,
Ryan

0 Jack Loop over 7 years ago in reply to Ryan Brown1

Guru 17870 points

Hi. I wanted to check in and see when the 3.0.2 version may come out?

0 Ryan Brown1 over 7 years ago in reply to Jack Loop

TI__Guru**** 217467 points

Hi Jack,

I apologize as there is a delay while we wait for stack certification.

Regards,
Ryan

0 YiKai Chen over 7 years ago in reply to Ryan Brown1

Guru 735685 points

Hi Ryan, do you have a new estimate release date?

0 Ryan Brown1 over 7 years ago in reply to YiKai Chen

TI__Guru**** 217467 points

I unfortunately do not, YK. Please stay tuned for an announcement.

Regards,
Ryan

Zigbee & Thread

Zigbee & Thread forum

CC2530: At 2v, CC2530 factory resets