13-Bit SADC

The Sense Analog-to-Digital Converter (SADC) is an ADC based on successive approximation registers (SARs). It is an 8-sensing-channel 13-bit 1 Msps ADC that can be used for single ended or for differential measurements. The SADC supports DMA to allow for large sample collection without overloading the MCU.

• 13-bit analog-to-digital conversion
• 8-channel (MSIO_0–MSIO_7) analog-to-digital conversion
• Up to 16 MHz ADC clock frequency
• Up to 1 Msps conversion rate
• Programmable ADC sampling clock
• Single-ended or differential measurements
• Internal temperature and Vbattery sensors
• Ability to capture ADC samples using DMA, unburdening the MCU.
• Programmable internal reference

ADC and Comp can share the input channel from MSIO, but MSIO selected for Comp cannot be used at the same time for ADC sampling. The selection and configuration of IO can be found in the Pin Mux and I/O section.

The SADC can sample one of 8 analog sources at a given time, using an input mux. Another two of the channel_N are used internally for VBAT and temperature monitoring for adjusting the analog and RF circuitry configurations for optimum performance. SADC can be reconfigured to perform single-ended or differential measurements through CFG.MODE register. The reference voltage can be internally generated or supplied as one of the external inputs and can be selected through CFG.REF_SEL register. The external reference source can be input through MSIO_0–MSIO_3, and the input voltage must not exceed one half of the supply voltage. The internal reference voltage can be programmed as 0.8 V, 1.2 V, 1.6 V, and 2.0 V through CFG.BIAS_RES_CTRL register. Affected by ambient temperature and chip manufacturing process, the internal reference voltage will have a maximum error of 20%. For any specific reference value (Vref), the maximum ADC input is supposed to be (2 x Vref). As shown in the following figure, the Iref comes from the relatively constant current generated by the chip's bandgap module and will not be affected by other modules; it will generate different Vref levels by resistive voltage division.

The ADC sampling clock can be configured as 16 kHz, 8 kHz, 4 kHz, 16 MHz, 8 MHz, 4 MHz, or 1 MHz through CLK.CLK_SEL register. The output value of the ADC will be stored into the FIFO, and users can transfer the data to RAM through DMA or MCU.

Due to the offset of the ADC itself, the original value of 0–8191 cannot accurately correspond to the voltage of 0–2*Vref, so the ADC of each chip needs to be calibrated. Usually, the slope and offset of the straight line are obtained by the method of linear fitting, and the current voltage can be converted by substituting the raw data.

• In the final test stage, the slope_int_0p8 and offset_int_0p8 of the 0.8 V internal reference voltage are obtained by measuring the two input voltages and their corresponding raw data, and the slope_int_0p8 and offset_int_0p8 are stored in the eFuse (refer to eFuse). You can use the following formula to convert the raw data into the current voltage value.

  $$\mathrm{cur}\_\mathrm{voltage}=(\mathrm{raw}\_\mathrm{data}-\mathrm{offset}\_\mathrm{int}\_\mathrm{0p8})/\mathrm{slope}\_\mathrm{int}\_\mathrm{0p8}$$

  Similarly, eFuse also stores slope_int_1p2, offset_int_1p2 of the 1.2 V internal reference voltage, and slope_int_1p6, offset_int_1p6 of the 1.6 V internal reference voltage. When the internal 1.2 V reference voltage is selected, the raw data can be converted into the current voltage value by the following formula:

  $$\mathrm{cur}\_\mathrm{voltage}=(\mathrm{raw}\_\mathrm{data}-\mathrm{offset}\_\mathrm{int}\_\mathrm{1p2})/\mathrm{slope}\_\mathrm{int}\_\mathrm{1p2}$$

  When the internal 1.6 V reference voltage is selected, the raw data can be converted into the current voltage value by the following formula:

  $$\mathrm{cur}\_\mathrm{voltage}=(\mathrm{raw}\_\mathrm{data}-\mathrm{offset}\_\mathrm{int}\_\mathrm{lp6})/\mathrm{slope}\_\mathrm{int}\_\mathrm{lp6}$$
• The slope_ext_1p0 and offset_ext_1p0 of the external 1.0 V reference source are also stored in the eFuse. When the external reference voltage is selected, you only need to pay attention to the scaling of the slope_ext_1p0. The conversion formula is as follows:
  $$\mathrm{cur}\_\mathrm{voltage}=\left(\mathrm{raw}\_\mathrm{data}–\mathrm{offset}\_\mathrm{ext}\_\mathrm{1p0}\right)/\left(\mathrm{slope}\_\mathrm{ext}\_1\mathrm{p0}*1.0/\mathrm{external}\_\mathrm{reference}\_\mathrm{voltage}\right)$$

The above conversion formulas are used when the single-ended mode is selected. If the differential mode is selected, the slope here will be reduced to one-half.

There is a voltage divider resistor inside the GR5526 to divide the voltage of the battery as VBATL/4. Users need to configure the CFG.CH_N register as 0xE (battery sensor as input), and select the internal reference voltage of 0.8 V.

Therefore, users can obtain the raw data of the current voltage, and then convert the raw data into the actual battery voltage value through the following formula:

Where, slope_int_0p8 and offset_int_0p8 are calibration offset and slope when 0.8 V internal reference voltage is selected; 3.854 represents the divider ratio of the resistor to the battery voltage. With the change of temperature, this divider ratio will be slightly different, ranging from 3.85 to 3.855; the typical value is 3.854.

GR5526 has a temperature sensor module inside, and the measurement range is from -40°C to 80°C. The temperature sensor converts the internal temperature of the chip into a voltage value as Vtemp and outputs it to the ADC for sample and conversion. After temperature calibration, the ADC converts the raw data into an actual temperature value.

If users want to measure the internal temperature of the chip, it is necessary to configure the CFG.CH_N register as 0xD (temperature sensor as input), and select the internal reference voltage of 0.8 V. The calibration formula is as follows:

Where, temp_25 represents an ambient temperature of 25°C; temp_25_offset represents ADC raw data at 25°C ambient temperature; slope_int_0p8 represents the calibrated slope value of the internal 0.8 V reference voltage; -0.00175 represents the amplification factor for the slope when used to measure temperature. In general, the output voltage decreases by a typical value of 1.863 mV for every 1 degree increase in temperature. The typical value of the voltage is 807.5 mV at -40°C and 583.9 mV at 80°C.

Although we have provided some calibration values stored in the eFuse at 3.3 V supply and normal temperature, these calibration values may have errors at other ambient temperatures. Users can use the following methods to recalibrate the conversion relationship between raw data and actual temperature or voltage.

• Single-ended mode calibration with internal reference source
  1. Select internal reference source by configuring the CFG.REF_SEL register, select internal reference level by configuring the CFG.BAIS_RES_CTRL register, and select single-ended mode by configuring the CFG.MODE register to initialize ADC.
  2. Use a noise-free power supply to connect to the N terminal by configuring the CFG.CH_N register to start ADC sampling.
  3. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  4. At this time, the average value is recorded as x0, and the input voltage is recorded as y0.
  5. Collect another voltage value using the above method, record the average code value as x1, and record the input voltage value as y1.
  6. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, y0, x1, y1 to obtain the offset and slope.
  7. The raw data can be converted into the actual voltage using the following formula:
  $$actual\text{\_}voltage=\left(raw\text{\_}data-offset\right)/slope$$
• Differential-ended mode calibration with internal reference source
  1. Select internal reference source by configuring the CFG.REF_SEL register, select internal reference level by configuring the CFG.BAIS_RES_CTRL register, and select differential-ended mode by configuring the CFG.MODE register to initialize ADC.
  2. Use two noise-free power supplies to connect to the P terminal and N terminal by configuring the CFG.CH_P register and the CFG.CH_N register to start ADC sampling.
  3. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  4. At this time, the average value is recorded as x0, and the value of the P terminal voltage minus the N terminal voltage is recorded as y0.
  5. Collect another voltage value using the above method, record the average code value as x1, and record the input voltage difference as y1.
  6. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, y0, x1, y1 to obtain the offset and slope.
  7. The raw data can be converted into the actual voltage using the following formula:
  $$actual\text{\_}voltage=\left(raw\text{\_}data-offset\right)/slope$$
• Single-ended mode calibration with external reference source
  1. Select external reference source by configuring the CFG.REF_SEL register, and select single-ended mode by configuring the CFG.MODE register to initialize ADC.
  2. Use a noise-free power supply to connect to the N terminal by configuring the CFG.CH_N register to start ADC sampling.
  3. Use a noise-free power supply to connect the ADC external input pin, and record the external input as z at this time (note that the external input is only connected from MSIO_0–MSIO_3).
  4. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  5. At this time, the average value is recorded as x0, and the input voltage is recorded as y0.
  6. Collect another voltage value using the above method, record the average code value as x1, and record the input voltage value as y1.
  7. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, y0, x1, y1 to obtain the offset and slope.
  8. The raw_data can be converted into the actual voltage using the following formula:
     $$actual\text{\_}voltage=\left(raw\text{\_}data-offset\right)/\left(slope*z/vref\right)$$

     where vref represents the voltage value of the external input as this time.

• Differential-ended mode calibration with external reference source
  1. Select external reference source by configuring the CFG.REF_SEL register, and select differential-ended mode by configuring the CFG.MODE register to initialize ADC.
  2. Use two noise-free power supplies to connect to the P terminal and N terminal by configuring the CFG.CH_P register and the CFG.CH_N register to start ADC sampling.
  3. Use a noise-free power supply to connect the ADC external input pin, and record the external input as z at this time (note that the external input is only connected from MSIO_0–MSIO_3).
  4. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  5. At this time, the average value is recorded as x0, and the value of the P terminal voltage minus the N terminal voltage is recorded as y0.
  6. Collect another voltage value using the above method, record the average code value as x1, and record the input voltage difference as y1.
  7. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, y0, x1, y1 to obtain the offset and slope.
  8. The raw data can be converted into the actual voltage using the following formula:
     $$actual\text{\_}voltage=\left(raw\text{\_}data-offset\right)/\left(slope*z/vref\right)$$

     where vref represents the voltage value of the external input as this time.

• Temperature sensor calibration
  1. Select internal reference source by configuring the CFG.REF_SEL register, and select internal 0.8 V reference source by configuring the CFG.BAIS_RES_CTRL register.
  2. Select single-ended mode by configuring the CFG.MODE register, and select temperature as N-terminal input by configuring the CFG.CH_N register to initialize ADC.
  3. Put the chip into a constant temperature environment.
  4. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  5. At this time, the average value is recorded as x0, and the current temperature is recorded as t0.
  6. Put the chip into another constant temperature environment.
  7. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  8. At this time, the average value is recorded as x1, and the current temperature is recorded as t1.
  9. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, t0, x1, t1 to obtain the offset and slope.
  10. The raw data can be converted into the actual temperature using the following formula:
  $$actual\text{\_}temperature=\left(raw\text{\_}data-offset\right)/slope$$
• Battery voltage calibration
  1. Select internal reference source by configuring the CFG.REF_SEL register, and select internal 0.8 V reference source by configuring the CFG.BAIS_RES_CTRL register.
  2. Select single-ended mode by configuring the CFG.MODE register, and select battery as N-terminal input by configuring the CFG.CH_N register to initialize ADC.
  3. Use a regulated power supply to supply power to the chip. Note that the power supply voltage of the chip that the ADC can work normally is 2.2 V–3.8 V.
  4. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  5. At this time, the average value is recorded as x0, and the current battery voltage is recorded as y0.
  6. Reselect another voltage value to power the chip.
  7. Sample 1024 pieces of data and remove the maximum and minimum values to obtain the average of 1022 numbers.
  8. At this time, the average value is recorded as x1, and the current battery voltage is recorded as y1.
  9. Use the formula $y=\left(x-offset\right)/slope$ , and bring in x0, y0, x1, y1 to obtain the offset and slope.
  10. The raw data can be converted into the actual battery voltage using the following formula:
      $$actual\text{\_}bat\text{\_}voltage=\left(raw\text{\_}data-offset\right)/slope$$

表 743 Electrical specifications

Parameter	Symbol	Min.	Typ.	Max.	Unit
Physical number of bits	PNOB		13		bit
Sampling frequency	Freq	4 K		16 M	Hz
Integrated nonlinearity	INL	-8		8	LSB
Differential nonlinearity	DNL	-8		8	LSB
Signal-to-noise and distortion ratio	SNDR		60		dB
Total harmonic distortion	THD		65		dB
Effective number of bits	ENOB		9.5		bit
Differential offset voltage, 13-bit resolution	Vos	-4		4	LSB
Gain error	Eg	-1		1	%
Sampling capacitor	Cs		2		pF
Sampling rate	F_S	0.27		1067	ksps
Input resistance	Rin		> 1 M		ohm
Current consumption during ADC conversion @100 Ksps	I_ADC		< 20		μA
Spurious free dynamic range	SFDR		68		dB