nhs3152 adc problem

[xiangxuejiao00]

dear sir 

   i encounter a question about test the battery voltage with the ADC function .

1. my schematic diagram is below 

2 my source about test the vcc voltage code is below 

static uint16_t Get_current_adc(void)
{
//int adc_max=0xfff;
int adcInput=0;
int total_adc=0;
//! [adcdac_nss_example_3]
Chip_IOCON_SetPinConfig(NSS_IOCON, IOCON_ANA0_4, IOCON_FUNC_1);
Chip_ADCDAC_Init(NSS_ADCDAC0);
Chip_ADCDAC_SetMuxADC(NSS_ADCDAC0, ADCDAC_IO_ANA0_4);
Chip_ADCDAC_SetInputRangeADC(NSS_ADCDAC0, ADCDAC_INPUTRANGE_WIDE);
Chip_ADCDAC_SetModeADC(NSS_ADCDAC0, ADCDAC_CONTINUOUS);
Chip_ADCDAC_StartADC(NSS_ADCDAC0);

int max,min;
total_adc=max=min=0;
for(int i=0;i<10;i++)
{
adcInput = Chip_ADCDAC_GetValueADC(NSS_ADCDAC0);
total_adc+=adcInput;
if(max<adcInput)
max=adcInput;
else if(min>adcInput)
min=adcInput;
}
Chip_ADCDAC_StopADC(NSS_ADCDAC0);
total_adc=total_adc-max-min;

//get the average value 

total_adc=total_adc/8;

//get the caculate from https://community.nxp.com/message/1119718#comment-1120963 
//output in mV = ((native value - native offset) * internal operating voltage / steps per uV) + offset
//native offset=2048,internal voltage = 1.2V,gain (steps per uV) = 2730,offset = 0.9V
adcInput=(total_adc-2048)*1200/2730+900;// get the test point voltage ,unit is mv

//convert the test point voltage to VCC voltage 
adcInput=adcInput*5/2;
return (uint16_t)adcInput;
/* Further handling of the threshold breach. */
}

my question is :

• how can i get these value ?i get from this value in another question ,but i do not no why ?i can not find it in spec ...native offset=2048,internal voltage = 1.2V,gain (steps per uV) = 2730,offset = 0.9V
• my ADC value adcInput  change rapidly,it is not a stable value ,

for example ,i get the voltage VCC by multimeter is 3.0V

but ,i tried 10 times test the value ,and get the adcInput = Chip_ADCDAC_GetValueADC(NSS_ADCDAC0);read from the register is below .

Header 1	Header 1	Header 2	Header 3	Header 4	Header 5	Header 6	Header 7	Header 8	Header 9
adcInput(mv)	2695	2677	2573	2596	2578	2676	2597	2670	2630
VCC(mv)	2960	2941	2826	2852	2832	2940	2853	2933	2889

can you give me some advise ?is there anything wrong with my schematic ?

is there anything wrong with my design formula ?

[driesmoors]

Hi,

It's a bit confusing to use a NHS3100 symbol in your schematics. - a good thing the digital pins are identical to the NHS3152 :-)

When the debug connector is still attached, a lot of noise is injected. Is it possible you measured those numbers within a debug session? If so, can you repeat your measurements without a debug session and without the LPC-Link2 debugger board connected to the NHS3152 IC?

Your fitting parameters used in the formula adcInput=(total_adc-2048)*1200/2730+900 do not ring a bell for me. Can you try to recall how/when/where you received those?

After the weekend, I will try and see if I can find other fitting parameters - we surely must have done this before, I need to check if I can find a trace of that somewhere...

Kind regards,
Dries.

[xiangxuejiao00]

dear dries

      thank you for your reply .

i am sorry that i make a mistake in the schematic ,i use 3152 not 3100 in my test .

1. the data ADC value that i test is not in debug mode without LPC-LIN2,i use CR2032 instead .
2. is there anything wrong in my bleeder circuit ?

3. the formula adcInput=(total_adc-2048)*1200/2730+900

i get the formula from https://community.nxp.com/message/1119718#comment-1120963 this document that you replied before 

1. can you give me some advise on test ADC value or any demo design ?

yours 

xuejiao 

[driesmoors]

Hi,

The formula given was derived from an early sample, and does not accurately reflect the state of the released ICs.

This formula provides a better conversion for production ICs:

adc_input_voltage = (native_value * 1.2V) / 2825 + 0.09V

(Of course, for your schematics, you need to retain your multiplier 2.5 as well).

Regarding reference design: we have a therapy adherence board available in our sample store, which uses resistive measurements (a.o. using the ADC) to keep track of pill intakes. The full schematics and layout is available in our SDK under <SDK>/docs/NHS3152 Therapy Adherence.zip

More information here: https://www.nxp.com/pages/:NHS3152THADADK

Kind regards,
Dries.

[xiangxuejiao00]

dear sir

i am sorry to trouble you again !

first :

i still confused about the formula  you offered before ,how do you get the formula ?

adc_input_voltage = (native_value * 1.2V) / 2825 + 0.09V

how to use reference voltage /  ADC resolution 12 bit and input range get the formula ?

second :

i have another question about ADC like before ,my schematic is still the picture like before ,i get the ADC raw  value from the register now ,and calculate the value average about 10 times .but there is still  about  15 --20 raw data fluctuation . i have no idea how to deal with the  fluctuation ,can you give me some advise ?

[cpi-ugo]

Could you please explain how this formula was derived from the NHS3152 datasheets since I believe it should be as follows 

adcInput = ((float)Chip_ADCDAC_GetValueADC(NSS_ADCDAC0) *
1600) / 4096;

[driesmoors]

Hi,

1

The ADC is not linear at the extremities, causing a slightly steeper slope of the conversion from raw to voltage. The formula takes this into account.

2

A 20 raw data fluctuation corresponds to ~8.5 mV, which is not that much. There is noise on the voltage input, and the IC being active also causes small fluctuations. You can use a scope to measure more precisely the voltage, as a multimeter is only showing averages. Perhaps best is to configure the scope in AC-coupled input mode to measure the ripple on the battery directly. You can also bring out a signal to synchronize the scope measurements with the ADC conversions.

Best,
Dries.

[xiangxuejiao00]

can you help me  ?driesmoorsKan