CC2340R5: Current spikes at intervals

Hi Shuyang,

Thank you for reaching out. Are there any modifications present in the basic_ble project? Is the device in a connection or simply advertising?

Best Regards,

Jan

Hello Shuyang,

The spikes appear to be pulses of recharge current, behaviour that is normal in this RF system, although they appear at a nearly constant frequency.

It would be helpful if you could give us more details on how you configured the basic_ble project (I assume in peripheral mode), such as the advertising parameters and if there is a connection established (although it seems there isn't). It is not possible to obtain much information from images at that scale.

Sincerely,

David.

Hi Jan, David, 

Thanks for the reply. I just changed the broadcasting interval to 1000ms and disable the UART menu in SysConfig, no other changes.

The spikes happens not regularly, my customer who reported this issue said it happened every 80-90 seconds, but I used EnergyTrace to take a 300s measurement, the issue was not captured every time, it took several measurements to reproduce it. If you have a power analyzer, it will be easier to see the issue.

I doubt if this is a recharge because the interval was too large? However I did not see recharge current like on Agama devices, I wonder if CC2340 has different recharge logic than Agama?

Best regards,

Shuyang

Hello Shuyang,

Could you please check (on the device package) if this is PG1.0 or PG2.0 (Silicon Revision)?

Silicon Revision

Best regards,

David.

It’s PG2.0

BR,

Shuyang

Hello,

Could you please send me the image that you are running on the device to reproduce it.

I modified the basic_ble project following your parameters but couldn't replicate the issue.

Best regards,

David.

Hi David,

Please find my project from the link below:

https://tidrive.itg.ti.com/a/Y9abz_7Wm_xx1pvX/166547fa-1e45-4089-8f6b-754e9c526229?l

It takes several measurements to capture the behavior because it happens from time to time. Are you using a power analyzer or EnergyTrace? A power analyzer would be easier to use because it can record result for longer time. If you are using EnergyTrace, please try to take a few more measurements, I used 5min measurement period and it took me 3-4 measurements to see the issue.

Best regards,

Shuyang

Hello Shuyang,

I ran the project several times for 5 mins (execution time) using EnergyTrace and I can not reproduce the same behaviour. I am working now on doing this with the power analyzer but so far I get the same results.

Do you have another board to try the example with?

Best regards,

David

Hi David,

I tried another LaunchPad and it also has this problem. And my customer also confirmed they saw the issue on all of their boards, it is not an individual behavior.

It just takes time to reproduce, from my observation it takes more than 90 seconds to happen, almost randomly happened.

Best regards,

Shuyang

Hello Shuyang,

I am trying to replicate the case with two other devices and for a longer period of time. I will keep you posted on the results.

In the meantime, could you please tell me if this behaviour occurs in a productive environment or in a controlled one used for testing (shield box vs crowded BLE environment) to have more details on this subject.

Best regards,

David.

It’s in office environment, other BLE signal do exist in the environment.

BR,

Shuyang

Hello Shuyang,

I will run some additional tests (for a longer period).

Please give me a couple of days to get back to you.

Best regards,

David.

Hi David,

Thanks for the support! Please let me know if you need anything from my side.

Best regards,

Shuyang

Hello Shuyang,

There are a couple of things that I want to double check.

1. The tests are running on the launchpad or a custom board? Just to double check.
2. I have been able to reproduce something similar to what you presented. Could you please reproduce again the issue and share with me the current energy trace graph showing that specific section in more detail (more zoom)? To have a more clear signal waveform and compare it to mine please.
3. Have you tried establishing a connection? If yes, is this behaviour negatively impacting the connection?

Best regards,

David.

Hi David,

1. It can be reproduced on both a LaunchPad and a customer board. My screenshots are from a LaunchPad.

2. Please find the zoomed image as below, is it the same as the expected recharge waveform?

3. I did 10 tests with 30s measurement after connection and did not see the issue so far. The connection interval was 30ms decided by Lightblue on the mobile phone. I think it also worth mentioning that the issue also happened when the customer disabled the advertising in basic_ble. The waveform was similar with the one with advertising.

Best regards,

Shuyang

Hello Shuyang,

Thanks, yes, I have very similar results but it is quite hard to reproduce them (seems random).

Can you always reproduced it (25%, 50%, 75%, 100% of the time)? or is it a one time thing?

What do you mean by disabling the advertising in basic_ble? Could you please specify this new configuration?

We are already looking further into this, so more information from your side on how to reproduce it will be very useful.

Best regards,

David.

Hi David,

It seems random to me as well, but I just double checked with the customer, they are seeing the behavior every 900s. If you have a power analyzer which can capture a long measurement, it should be easier to see it.

By disabling the advertising I mean to comment out the GapAdv_enable call in the code.

The customer also said they saw it during a connection.

Best regards,

Shuyang

Hello Shuyang,

Please help me with the following:

1. Verify if this behaviour is also happening when the advertising interval is 100 ms or less.
2. Please confirm that there are no other peripherals connected and only BLE is being used.

Best regards,

David.

Hi David,

I just saw the release note of the new SDK 7.20.01.10 and there is a know issue which looks very similar to our issue:

"CC2340 unintended wakeup from standby, interval gets shorter and shorter"

https://sir.ext.ti.com/jira/browse/EXT_EP-11465

Would you please help double check if this is the same issue we are seeing?

Best regards,

Shuyang

Hello Shuyang,

Yes, this is indeed the same issue. We are currently working on it.

I will provide more details as soon as possible.

Best regards,

David.

Hello Shuyang,

The team has found the rout cause of the issue and it is working to solve it, already showing positive results at reducing the frequency of the events.

Could you please let me know if this is causing any issue in production that needs to be addressed urgently?

Best regards,

David.