Custom nRF52840 module high power consumption with OPENTHREAD turned on

[zhangxka]

I have a custom nRF52840 module (made by Holyiot 18010) running the default Blinky program with some custom configs. I'm measuring its power consumption with PPK II with source mode on a 3V voltage. If I run the program normally the current consumption hovers around ~10uA (which is to be expected). However, if I have CONFIG_NET_L2_OPENTHREAD=y, the average current jumps to 1mA and stays there. 

What is strange is that when I try to run the same program with OPENTHREAD turned on for the nrf52840 DK, the current seems to be back in the uA range. Any idea what could be the problem here? Appreciate any help!

[Marte Myrvold]

Hi,

Does the module have an external 32 KHz XTAL? If not, please try enabling the RC oscillator in your prj.conf:

CONFIG_CLOCK_CONTROL_NRF_K32SRC_RC=y

Best regards,
Marte

[zhangxka]

Hi Marte,

Thanks for your reply! It does look like it has an external 32 KHz XTAL. Please see the schematics of the custom module below:

[zhangxka]

Hi Marte,

Is there another reason you might think of why this is happening? Setting CONFIG_CLOCK_CONTROL_NRF_K32SRC_RC didn't seem to help. Thanks again!

[Marte Myrvold]

Hi,

Have you configured the DC-DC?

What is strange is that when I try to run the same program with OPENTHREAD turned on for the nrf52840 DK, the current seems to be back in the uA range

Can you specify what the current is on the DK?

Have you enabled sleepy end device on the nRF52840 module, and is what you are measuring the sleep current or when it is doing something?

Best regards,
Marte

[zhangxka]

Hi Marte,

I've tried to cut down the options and the code to a minimum. The options in my prj.conf file are

CONFIG_SERIAL=n

CONFIG_NETWORKING=y

CONFIG_NET_L2_OPENTHREAD=y

(Run on DK about 10uA; on custom module about 1mA)

If I set CONFIG_NET_L2_OPENTHREAD=n instead, it will give me about the same power usage on DK (about 10uA), but I get much lower current consumption on my custom board that lowered to about ~30uA.

Running the boards as SED devices didn't really make a difference in this minimal example for me since I'm assuming the board is not making any active connections.

What's really bizarre is that if I supply the custom board with a much lower voltage on nRF_VDD (the cutoff seems to be about 2.5V), it seemed to have triggered something off. The voltage on the custom board is now running in the ~50-70uA range. This will only happen after I do a reset after the voltage set. Please see the following graph for reference.

Best,

Kevin

[haakonsh]

Hey Kevin,

Your VDDH net is floating. When VDDH/REG0 is enabled, left floating, and more than ~2.5V is supplied to VDD, then current will flow back into VDDH/REG0 and REG0 enters an undefined state.

See Circuit configuration no. 4 for QIAA aQFN73 for reference.

BR,

Håkon Holdhus.

[zhangxka]

Thank you!! That does seem to be the issue. Is there any way to turn VDDH/REG0 off through config options or firmware since I don't have easy control over the how the custom module was designed?

[haakonsh]

Set BOARD_ENABLE_DCDC_HV=n in your prj.conf file. You also need to remove L4 and C19, and connect VDDH to VDD with a short wire. That should put the board in a good state, though there might be permanent damage to the REG0 stage. Also note that max input voltage is now 3.9V instead of 5.5V. 

BTW, was this module designed for you by HolyIOT or is this one of their off-the-shelf modules? 

[zhangxka]

Hey Håkon, unfortunately this is one of their off the shelf modules. Would source another nRF module my only other option then for a low power consumption IOT device? It seem to be a design error on the HolyIOT part and there’s no way to override their module connections 

[haakonsh]

Yes, you need to source another module. See our list of Third-party modules. I've notified HolyIOT about the issue, but I don't know if they will fix or how much time it will take. You can contact them directly and ask them yourself. They will probably fix the issue and sell you their new revision directly if they are able to.

[zhangxka]

Sounds good. Thanks for the help!

[zhangxka]

Hey Håkon, 

I have decided to use another module (E73 for now). I have attached their schematics below and modified the external circuit using sample circuit #2 as a reference. Can you help me take a look and make sure the wires/components I added make sense? 

One thing to note here is that I am not using VUSB to supply any powers. It's there to detect the presence of USB power source so that I get the charging event through USB module.

Best,

Kevin