Dear Support Team,
We have developed a product based on the nRF5340, using the Insight SIP ISP2053 module.
So far, we have built approximately 25 prototype units and around 50 boards from our pilot production run. For functional testing, we use our own firmware, which is based on nRF Connect SDK 3.3.4. Various board tests are executed via shell commands. The shell interface is configured on UART0, and we use the SEGGER J-Link Virtual COM Port to connect this UART interface to the test PC.
We have observed that all prototype boards draw approximately 500 µA (CPU idle, UART0 enabled and waiting for data), while the boards from the pilot production run draw approximately 900 µA.
The boards are assembled identically; the only difference is that the ISP2053 modules (and so the nRF5340) originate from a different production batch. The exact same firmware is programmed onto both board types.
The part identification values in the FICR are the same for both boards:
INFO.PART = 00005340
INFO.VARIANT = 434C4141
INFO.PACKAGE = 00002000
We have verified that the UART register and clock config registers contents are identical on both board variants while the higher current consumption is present.
UART0.ENABLE = 0x00000008
UART0.CONFIG = 0x00000000
UART0.BAUDRATE = 0x01d60000
UART0.PSEL.TXD = 0x00000019
UART0.PSE.RXD = 0x0000001b
UART0.PSEL.RTS = 0xffffffff
UART0.PSEL.CTS = 0xffffffff
UART0.INTEN = 0x00000010
UART0.RXSTARTED= 0x00000001
HFCLKRUN = 0x00000000
HFCLKSTAT = 0x00000000
LFCLKRUN = 0x00000001
LFCLKSTAT = 0x00010002
HFCLKAUDIORUN = 0x00000000
HFCLKAUDIOSTAT= 0x00000000
CLLK192MRUN = 0x00000000
CLKK192MSTAT = 0x00000000
HFCLKSRC = 0x00000001
LFCLKSRC = 0x00000002
HFCLKCTRL = 0x00000001
When we disable UART0 on both boards, the current consumption drops to approximately 20 µA in both cases. Therefore, we can largely rule out assembly issues or incorrectly configured I/O pins. The following code causes the current consumption on both board versions to drop to approximately 20 µA:
pm_device_action_run(DEVICE_DT_GET(DT_NODELABEL(uart0)), PM_DEVICE_ACTION_SUSPEND);
The higher current consumption does not negatively affect our application, but we are concerned because we cannot explain the difference.
Is it possible that the current consumption of the nRF5340 device can vary this significantly between different production batches, or is there another explanation that you would consider more likely?
Thank you for your support.
Kind regards,
Ueli Hanselmann