jaredwolff yes, we have an external antenna, for example mounted on the top of a car rather than inside the steel box where the unit is connected to power.
nRF9160 Feather - Suggestions for next version
jaredwolff
Hi Jared,
Like the majority of the “makers”, nearly all the sensors I use are 3.3V. However, for the longer term, I think it’s worth having a configurable 1.8-3.3V VDDIO if not much additional effort.
Are you planning on driving the RP2040 at 3.3V or in addition also thinking of using 1.8V, as RP2040 does support multiple power options?
Will there be an option to shutdown RP2040 if nRF9161 is in a sleep mode?
Any plans to bring out any of the GPIOs from the RP2040 to the Feather pins?
- Edited
@zpm1066 thank you for the amazing feedback! 
zpm1066 Like the majority of the “makers”, nearly all the sensors I use are 3.3V. However, for the longer term, I think it’s worth having a configurable 1.8-3.3V VDDIO if not much additional effort.
It is possible actually with the nPM1300. The only part that wouldn’t work is the onboard NOR flash. It does not support such a wide voltage range. They have 1.8V variants and 3.3V variants. My thoughts would be to go with the 3.3V version and if any flash writes need to occur the buck needs to be placed into 3.3V mode. Testing is still necessary since I’m not sure how the flash would handle 1.8V. If it ends up leaking tons of current that would put the kibosh on it right there.
zpm1066 Are you planning on driving the RP2040 at 3.3V or in addition also thinking of using 1.8V, as RP2040 does support multiple power options?
Right now, thanks to all the feedback thus far, the plan is to run it at 3.3V. It does have a separate VDDIO pin which allows for running the IO at 1.8V. It does require 3.3V for USB (which is critical for this board!) on the other power supply pins no matter what.
zpm1066 Will there be an option to shutdown RP2040 if nRF9161 is in a sleep mode?
Good questions. It’s the reason why I’m asking you guys your preferences on IO voltages. 
By default it’s on on a cold boot but then is configurable in software to be shutdown when not in use. Both buck converters actually can be configured in “retention mode” which lowers the rail voltages when in a sleep mode (very neat).
zpm1066 Any plans to bring out any of the GPIOs from the RP2040 to the Feather pins?
No plans since it’s mostly meant to be a programming IC. All the GPIOs are going directly to the nRF9161.
jaredwolff By default it’s on on a cold boot but then is configurable in software to be shutdown when not in use. Both buck converters actually can be configured in “retention mode” which lowers the rail voltages when in a sleep mode (very neat).
Wonderful! Your nRF9161 Feather is looking fantastic!
Btw - Any decision on which RP2040 debug probe firmware (probe-rs or raspberrypi debugprobe) you’re leaning towards? probe-rs development seems to be pretty active. I haven’t yet tried either but plan to in the next month or so.
zpm1066 no choice yet. The CMSIS-DAP C implementation is slightly faster. I haven’t gotten there yet though. Stay tuned 
I guess, it’s a too specific use-case, but I would welcome, if the board offers a 8-pin EEPROM solder mask. I would equip that with an HSM (ATECC608A) and use that for the encryption with tinydtls.
I know, that the nRF9161 offers TF-M as well and so a HSM may be considered to be obsolete, but for portability it’s easier to keep the keys in such a device as the ATECC608A.
@jaredwolff
Perhaps a little late but is there any chance of adding an I2C/STEMMA connector for sensors, OLED, EEPROM, etc? Not sure if there is any space on the reverse side of the board. It would be pretty useful for expansion. Thanks.
Thanks for the suggestions. I’ll definitely keep them in mind.
16 days later
- Edited
AchimKraus Thingy:91 is missing expandability and a DK is just to large in size.
Recently Nordic Semi has been making a few changes to the Thingy91:X devicetree files, e.g. added support for the expansion board and other misc changes. I don’t see an “edge_conenctor” defined in the "“thingy91x_nrf9151” devicetree yet like in nRF5340. Somewhat slow progress at the moment, perhaps there are delays in nRF9151 availability.
I expect the Thingy91:X Expansion Board will be the same spec as that in the Thingy:53 (nRF5340) or an enhanced version with backward compatibility. The nRF7002-EB WiFi 6 ( I use several with Thingy:53 and nRF52840 boards) should work with the Thingy:91X. With I2C (EB and I2C STEMMA), SPI and several GPIOs, the Thingy:91X should be pretty expandable for sensors and peripherals.
Originally, both the Thingy91:X and nRF9131-EK were planned for Q1 release. Hopefully we’ll see a Q3 release.
Yea it will be cool to see. Who knows I may base a new board off of the 9151 
First thing is first, gotta finish the 9161 version!
a month later
Proto 2 has been on my desk for a few days now. I’m trying to figure out what’s drawing so much current with the programmer disabled. (Getting 1mA in active sleep whereas it should be < 10uA)
I think it’s going to have to come to removing parts until I get the expected current. 
We’ll see!
- Edited
jaredwolff Looks great! What’s the beige connector? I2C STEMMA?
zpm1066 jaredwolff Looks great! What’s the beige connector? I2C STEMMA?
Currently connected to the SWD lines for the RP2040 for programming with an external Picoprobe
- Edited
Now we’re talking!
Average of 6uA in sleep mode at 3.6V. That includes the LIS2DH and the external flash in sleep/power down. Slightly better than the current nRF9160 Feather at ~10uA.
jaredwolff Hi Jared, Awesome!
What was driving the current high?
One culprit is the UART console. That should be disabled when the device is operating in low current applications:
int setup_uart()
{
static const struct device *const console_dev =
DEVICE_DT_GET(DT_CHOSEN(zephyr_console));
/* Disable console UART */
int err = pm_device_action_run(console_dev, PM_DEVICE_ACTION_SUSPEND);
if (err < 0)
{
printk("Unable to suspend console UART. (err: %d)\n", err);
return err;
}
/* Turn off to save power */
NRF_CLOCK->TASKS_HFCLKSTOP = 1;
return 0;
}I was having issues with NCS 2.5.x so when I switched to 2.6.1 I had much better luck configuring the nPM1300. I’m aiming to release the board synced with NCS 2.7.x.
jaredwolff Nice! Makes sense that the UART be disabled for low current usage.
I’m glad that the nRF9161 board will be released with NCS 2.7.x. Latter has some major changes in SoC / board names and definitions, plus Sysbuild for multi-image/multi-SoC support.
I have NCS 2.7.0 (pre-release) installed but will wait for a “release Version” before moving a few of the multi-image builds to Sysbuild.
Is the beige connector going to be for I2C STEMMA? Thanks.
zpm1066 Is the beige connector going to be for I2C STEMMA? Thanks.
Currently connected to the SWD lines for the RP2040 for programming with an external Picoprobe
jaredwolff
Is the plan to let it stay connected to SWD on board release or use it for I2C and program the board via USB? Thanks.
- Edited
The RP2040 can be programmed with the built in bootloader. The nRF9161 is programmed with the RP2040 running CMSIS DAP.
I actually don’t necessarily need the extra connector for programming purposes. Something like Qwiic or STEMMA would be attractive for you @zpm1066?