Current draw significantly higher than in the product specification (nrf52840)

RE: Current draw significantly higher than in the product specification (nrf52840)

lx_brz — Mon, 04 Mar 2024 13:33:29 GMT

Hi Kenneth,

thank you very much for checking and confirming that probably this is not an error on my side. I'll mark your last comment as "verify answer".

Best regards

Alex

RE: Current draw significantly higher than in the product specification (nrf52840)

Kenneth — Mon, 04 Mar 2024 13:08:19 GMT

Hi Alex,

Looking at the distribution curve I can see that most samples are within 2-3uA, but 5.2uA is within prcoess variation, though at the very high end of the variation. So this is close to the max you should expect yes.

Kenneth

RE: Current draw significantly higher than in the product specification (nrf52840)

lx_brz — Fri, 01 Mar 2024 08:32:57 GMT

Hi Kenneth,

> I think 5.2uA is still high, what is the system off current for this chip?

With a call to `sys_poweroff();` in zephyr I measure 517 nA. Seems to be much lower than the 0.97 uA specified in the product specification for "I ON_RAMOFF_EVENT"

> can you provide the detailed chip markings printed on top of the IC?

Sure: N52840 / FQAAF0 / 2311AD

> nrfjprog --readficr ficrdump.hex

devzone.nordicsemi.com/.../ficrdump_5F00_chip_5F00_5_5F00_point_5F00_2_5F00_uA.hex

Just for completeness, I did the same for the other chip that resulted in 3.6 uA measured:

sys poweroff: 400 nA

markings: N52840 / FQAAF0 / 2311AD (same as above)

ficr:

devzone.nordicsemi.com/.../ficrdump_5F00_chip_5F00_3_5F00_point_5F00_6_5F00_uA.hex

Thank your for your help.

Alex

RE: Current draw significantly higher than in the product specification (nrf52840)

Kenneth — Thu, 29 Feb 2024 15:25:42 GMT

Hi,

Also, for the 5.2uA chip, can you provide the detailed chip markings printed on top of the IC? Also, readout the ficr so I can check production data for the chip in question by: nrfjprog --readficr ficrdump.hex

Kenneth

RE: Current draw significantly higher than in the product specification (nrf52840)

Kenneth — Thu, 29 Feb 2024 15:22:26 GMT

Hi Alex,

Pin configuration looks good yes.

I think 5.2uA is still high, what is the system off current for this chip?

Kenneth

RE: Current draw significantly higher than in the product specification (nrf52840)

lx_brz — Thu, 29 Feb 2024 14:46:42 GMT

Hi Kenneth,

thank you for the quick reply. The pins seem to be configured correctly

P0: nrfjprog --memrd 0x50000700 --n 0x80

0x50000700: 00000002 00000002 00000002 00000002   |................|
0x50000710: 00000002 00000002 00000002 00000002   |................|
0x50000720: 00000002 00000002 00000002 00000002   |................|
0x50000730: 00000002 00000002 00000002 00000002   |................|
0x50000740: 00000002 00000002 00000002 00000002   |................|
0x50000750: 00000002 00000002 00000002 00000002   |................|
0x50000760: 00000002 00000002 00000002 00000002   |................|
0x50000770: 00000002 00000002 00000002 00000002   |................|

P1: nrfjprog --memrd 0x50000A00 --n 0x40

0x50000A00: 00000002 00000002 00000002 00000002   |................|
0x50000A10: 00000002 00000002 00000002 00000002   |................|
0x50000A20: 00000002 00000002 00000002 00000002   |................|
0x50000A30: 00000002 00000002 00000002 00000002   |................|

> Whether running code from Flash or RAM when going to sleep have no impact on current consumption. Neither does cache.

Ok, was not sure about this.

Interestingly I measured with a second board, which actually has some more components on it. (We prepared a set of boards each step-by-step less components to analyze current consumption caused by the individual components on the board). This board has the pull up resistors of the two I2C lines still soldered in. The pins are configured to a high signal:

0x50000700: 00000002 00000002 00000003 00000003   |................|
0x50000710: 00000002 00000002 00000002 00000002   |................|
0x50000720: 00000002 00000002 00000002 00000002   |................|
0x50000730: 00000002 00000002 00000003 00000002   |................|
0x50000740: 00000002 00000002 00000002 00000002   |................|
0x50000750: 00000002 00000002 00000002 00000002   |................|
0x50000760: 00000002 00000002 00000002 00000002   |................|
0x50000770: 00000003 00000003 00000003 00000002   |................|

Here I am measuring 3,6 uA. Interesting. So this is due to normal variation from chip to chip? This measurement and the previous one were done at room temperature.

Best regards

Alex

RE: Current draw significantly higher than in the product specification (nrf52840)

Kenneth — Thu, 29 Feb 2024 14:26:25 GMT

Hello,

Whether running code from Flash or RAM when going to sleep have no impact on current consumption. Neither does cache.

Even though there is some variation of current consumption from chip to chip, it sounds high to measure 5.2uA unless you are well above normal room temperature. I am thinking the problem here may be related to some pins are configured as input and floating, can you just check that none of the pins are configured as inputs and might be floating? For instance you can readout the pin configuration for port P0 and P1 and make sure all values are 0x2?

P0: nrfjprog --memrd 0x50000700 --n 0x80
P1: nrfjprog --memrd 0x50000A00 --n 0x40

Kenneth