nhs3100 current consume

xiangxuejiao00 · ‎09-11-2019

dear

recently ,i encounter a question about the 3100 current consume .

i use release_mra2_10_1_nhs3100 SDK packet , i use the default NHS3100 Temperature Logger project and just remove the define of BOARD_ENABLE_WAKEUP and set the PIO0.0 PIN pull up ,because i do not use the wake function .

if i connect the wake up pin to the VCC, i get the current consume is 10μA both in power off mode and dpd mode ;

but if i connect the wake up pin to the GND, i get the current consume is 10μA in power off mode ,and get the current consume is 50μA

in dpd mode .

i want to know is this correct result when correct the wake up pin to gnd?why this occurs ?

what can i do to decrease the current consume without change the pcba design ? i

xiangxuejiao00 · ‎09-15-2019

dear Dries

Here is my schematic

my purpose is wake up the 3100 through wake up pin when the lis3dh bigger than threshold .

i do the test as follow

test ：

set the 3100 wake up pin pull up ,do not set the lis3dh,do not set 3100 (so it is in power off mode ),now i get the whole current consume is 13.2 μA, it seems correct
then i config the 3100 work every 10 seconds ,and still do not care about the lis3dh,now 3100 is in deep power down mode if the RTC not come . now i get the whole current consume is 57.7μA,it seems not correct

in fact ,i just change the 3100 work or not ,but the current consume is so different as i expect .

is there anything wrong in my design ?or what should i do to config the 3100?

I am anxious to get your reply

BR,
xuejiao.

driesmoors · ‎09-16-2019

Hi,

Some remarks:

R2 is not needed, as the RESETN pin has a pull-up internally.
C4 is not needed.
You are powering the 3 ICs directly from the battery. The humidity sensor and accelerometer will draw current even when not in use. To have the lowest possible current consumption, you must provide current through the NHS3100, i.e. connect a GPIO pin to the VDD pins of the sensors.
Are you measuring current via a reserved pin? The reserved pins must be tied to ground. You can measure the current consumption by means of a resistor in series close to the battery.
Try to remove R5. Now voltage is pushed to the NHS3100, also in power-off or in DPD. This results in extra current consumption.

About the software:

In SDK 10.1, the PIOs are all set with internal pulldowns. Given your current schematics, you can reduce the current consumption by changing that to a pull-up for PIOs 0 and 8.
You can download the latest SDK from https://www.nxp.com/pages/:NHS3100 - we're now at SDK 12.1 BETA. Quite a few changes have been introduced, but also a lot of bug fixes. It is highly recommended to at minimum move on to SDK 11.2 stable.
In SDK 12, we introduced support for a new board, our NHS3100 Sensor Board. This is a board which combines the NHS3100 IC with the NXP accelerometer FXLS897X and the NXP UHF IC SL3S4011FHK, plus a third-party humidity sensor - should sound familiar ;-)
In the SDK, the full schematics and layout of this new board are available, plus a demo application with full sources, and APPs for all major platforms. A fact sheet is available here: https://www.nxp.com/docs/en/fact-sheet/NTAGSENSORFS.PDF May be worth a look?

Best,
Dries.

xiangxuejiao00 · ‎09-16-2019

dear Dries.

ok,i will have a try as you mentioned before .

by the way ,where can i get the SDK12 and new NHS3100 Sensor Board?

driesmoors · ‎09-17-2019

Hi,

Go to https://www.nxp.com/pages/:NHS3100 > TOOLS & SOFTWARE, then scroll down. Yu can find NHS3100 SDK - v12.1.1 BETA under "Embedded Software".

Best,
Dries.

xiangxuejiao00 · ‎09-17-2019

dear Dries.

i tried the remark you mentioned before ,and found the mistake i have made in the schematic ,i connect get the current consume as below .

question 1： i find NHS3100 SDK - v12.1.1 BETA ,but it just contain the doc folder ,there is no new sensor board or schematic in the folder .

question 2：

as you mentioned before :

To have the lowest possible current consumption, you must provide current through the NHS3100, i.e. connect a GPIO pin to the VDD pins of the sensors.

i search the port ,but only find one VDD pin,its function is just as input ,there is no othen output VDD pin can use .

if i use the gpio port as output VDD,how can i judge whether it is enough to driver the other sensor IC ?

I am anxious to get your reply

BR,
xuejiao.

driesmoors · ‎09-20-2019

Hi,

On the SKD download:

Thank you for pointing this out. The link points to the wrong file. I contacted the webteam and this will get fixed in the coming days.

On using the gpio port:

The voltage routed via the NHS will be a bit lower than when taken directly from the battery, as the pads have a resistance of ~25 to 30 ohm. You'll need to check the datasheets to see if your use case can live with this (small) loss.

The sensor board we are finalizing does this, and when using our FXLS8471Q: ±2g/±4g/±8g, Low g, 14-Bit Accelerometer | NXP the combination can work together. If you can bear with us while we fix the SDK download, you can have a look at how we tackled it in the schematics that are included (<SDK>/docs/NHS3100 SensorBoard.zip)

Kind regards,
Dries.

xiangxuejiao00 · ‎09-24-2019

dear Dries.

i am sorry to trouble you again ,i have another question about the current consume .

in the 3100 SDK package ,i find a tool called CurrentConsumption.xlsx, i can easily calculate the current consume and how long the 3100 can work .

but i want to use it in my 3152 project ,i am going to calculate how long the 3152 can work according to battery voltage get from adc pin ,my whole system current consume ,and work interval ,etc.

so ,i want to know how to calculate the result ?i cannot find the computational formula in the tool ,?can you offer it to me ?or can you give me other advise to calculate the life span of the battery ?

I am anxious for your reply

yours

xuejiao

driesmoors · ‎09-24-2019

Hello Xuejiao,

No problem.

The tool CurrentConsumption.xlsx assumes an execution flow as depicted in that tab, i.e. where the bulk of the lifetime is spent in Deep Power Down mode, with only short active periods, in which measurements are initiated, calculations are performed, and data is stored in NVM.
Assuming that in your use case, using the NHS3152, the execution flow is similar, the battery capacity estimations will likely also be similar. To check or change the numbers it is based upon, you can unprotect the workbook and the worksheets (there is no password set), then unhide a helper worksheet. You then have access to all data and formulas it is based upon.

The main giveaway from this tool is to convince yourself that the Deep Power Down mode is the mode that matters: doubling the active time duration as proposed in the tool, or changing the system clock frequency (which has a big impact on the active current consumption) hardly moves the calculated estimations. Of course, the battery must be able to cope with the maximum load during active periods, still providing a sufficiently high voltage under all circumstances. But to determine the capacity, you need to make sure that in your layout, current consumption is minimized while in Deep Power Down mode.

Kind regards,
Dries.

xiangxuejiao00 · ‎09-25-2019

dear Dries.

yes ,i can use this tool and get the result of the current consumption with different parameter .but what i want to know is how did this tool work .i also unprotect the workbook and the worksheets , then unhide a helper worksheet .but i still confused how does these parameter affect the result ?

for example ,

how this tool get the value of tlogger_current_day ?
what is the relation between tlogger_current_day and the parameter before ?

i cannot find formulas about the value of tlogger_current_day in this Excel .

driesmoors · ‎09-27-2019

Hi,

On tab TLOGGER, cell G30 is named cycle_time, cell H30 is named cycle_unit
cycle_time and cycle_unit are used using the INDEX and MATCH functions to get the correct factor from table A71:E75.
That factor is used on the hidden sheet in cell D40 (named tlogger_current_day) in combination with the sum on cell D39. The formula is visible if you highlight the cell D40.

The sum on D39 amasses the consumption of the different blocks and takes their respective durations into account. If you start the exercise from scratch, you should end up with the same numbers.

Regards,
Dries.

xiangxuejiao00 · ‎10-09-2019

dear Dries.

i am sorry to disturb you so much ,but i encounter another question about the current consumption .

i found that ,when the phone with nfc close to 3100，the current is 500 μA, i do not use any application ,just offer the nfc

field by the phone .

but ,if 3100 wake up by rtc ,the current is just no more than 100 μA

according to my understanding ,these two wake up procedure quiet simsimilar .and i can not find the nfc current consumption in spec

so ,i want to know why the current consumption is so big when the 3100 wake up by nfc?

what can i do to decrease the current consumption ?because i have to take this into consideration when calculate the system current

consumption .

i am looking forward to hearing from you !

thanks

Xuejiao

driesmoors · ‎10-11-2019

Hello Xuejiao,

500 μA is a very high number. We see a small increase when bringing in an NFC field: the NFC shared memory is then clocked at half the RFiD frequency (so, 6.78 MHz) instead of the ARM System clock frequency. This results in an increase of maximally ~63 μA (when jumping from 62.5 kHz to 6.78 MHz).

Can you try to isolate the problem? By disconnecting peripherals you may be able to find the issue.

That aside, the required battery capacity is not affected by a high current consumption during NFC sessions (mind, the high number you measured does indicate a problem): over the full lifetime of the solution, the time spent in an NFC field is extremely small compared to time spent when in Deep power down mode, for example.

Best,
Dries

xiangxuejiao00 · ‎10-17-2019

can you help me abount this question?https://community.nxp.com/thread/515267

xiangxuejiao00 · ‎09-22-2019

OK ,i am looking forward to the new SDK version ,

Thank you for your patient reply these days .

driesmoors · ‎09-12-2019

Hello 向雪姣

With a battery connected, the power-off current consumption can be as low as 30-50 nA, and the Deep power down current consumption 2-3 μA.

With you a pull-up on the wake-up and that same pin connected to ground, current will be continually leaked. While in deep power down the state of the pins is kept, I highly suspect this is the cause of your jump from 10μA to 50μA.

Can you share more details of your design? Either here, or in a private message? Depending on what is connected to the pins, you may need to adapt the configuration in firmware before going to deep power down and power-off.

BR,
Dries.