CN-0411

Circuit Note

Rev. 0 | Page 8 of 12

Auto-Ranging Conductivity Measurement

For a given type of solution and conductivity probe used, the

range of conductivity measurement range encompasses multiple

gain resistance settings. Furthermore, to prevent the excitation

voltage from exceeding the rated voltage of the conductivity

cell, the excitation voltage value must change along with the

gain resistance. The software is capable of automatically setting

the gain resistance and the excitation voltage to the range of the

conductivity of the solution. This ensures that the conductivity

reading is always being measured at the most reliable setting.

The method of auto-ranging the conductivity measurement is

described in Figure 12.

Figure 12. Auto-Ranging Conductivity Measurement Flowchart

Calibration Procedure

The system is capable of two calibration methods, a zero-scale

calibration and through a reference precision resistance. The

calibration method used to obtain the measurements discussed

in Figure 15 uses a reference precision resistance, which can

either be a 20 Ω or 200 Ω resistor as shown in Figure 5. It is in

these resistance ranges that the on-resistance of the multiplexer

has a significant effect on the conductivity measurement. The

output of this calibration is an offset resistance value, which is

added to the gain resistance used in the computation of

conductivity. Figure 13 shows the procedure to perform this

calibration.

Figure 13. Flowchart of Calibration Procedure through a Reference

Resistance

The zero-scale calibration performs the zero-scale calibration of

the AD7124-8, which eliminates the output bias voltages of the

sample-and-hold topology. This ensures that a zero-input to the

conductivity cell corresponds to the ADCs zero code. Place a

shunt in Pin 2 and Pin 3 of Jumper Header P6 of the CN-0411

board. The AD7124-8 is commanded to perform system zero-

scale calibration on both the positive and negative voltages of

the conductivity signal. This stores the zero-input voltage level

in the ADCs offset registers and is automatically considered in

every sample reading.

Both these two calibration methods need to be done only once

for each board since the software stores the output of the

calibration procedures.

System Accuracy

From Equation 9, the computation of conductivity depends on

the two output voltages of the sample-and-hold topology, RGAIN,

and the DAC output voltage. RGAIN resistors below the MΩ

range have 0.1% tolerance while the 2 MΩ and 20 MΩ resistors

have 1% tolerance. Added to the resistances in the simple

voltage divider network shown in Figure 3 are the on-

resistances of the multiplexer ADG1608 and the conductivity

cell switch ADG884, which are maximum at 17.4 Ω and 0.96 Ω,

respectively. The input bias currents and input offset current

from the instrumentation amplifier introduces a voltage, which

scales with the gain resistance and the resistance of the solution.

The B-grade AD8220 has a max input bias current of 10 pA for

each input and a max input offset current of 0.6 pA, which

makes a total input bias current of 20.6 pA. At 20 MΩ, this

constitutes a 20.6 pA × 20 MΩ = 412 µV input bias voltage.

Moreover, at a gain of 10, the AD8220 has a 0.2% max gain

error. Computing conductivity directly from a known precision

resistance measures the system accuracy and accounts for all

these factors, including the ones introduced by the sample-and-

hold topology, such as the drop voltage in Equation 4.

START

SET MULTIPLEXER

TO HIGHEST

GAIN RESISTANCE

(20MΩ)

READ ADC INPUT

V+ AND V–

YES

NO

PERFORM CONDUCTIVITY

MEASUREMENT AND

TEMPERATURE

MEASUREMENT

CALCULATE TDS

DROP GAIN

RESISTANCE

BY 1 DECADE

V+ + V– > 0.3 × 2VEXC

SET DAC TO

OUTPUT VEXC = 400mV

(OR SPECIFIED USER

DAC VOLTAGE VDAC)

SET VEXC =

VDEF × VEXC

V+ + V–

START

SHORT TERMINALS

3 AND 4, OR 5 AND 6,

IN JUMPER P5

PERFORM ADC SAMPLING

AND COMPUTE V+ AND V–

FROM THE INPUT

POSITIVE AND NEGATIVE VOLTAGES

COMPUTE OFFSET RESISTANCE

ROFFSET =

WHERE:

RREF IS THE REFERENCE PRECISION RESISTOR TOLERANCE

VEXC IS THE CURRENT EXCITATION VOLTAGE VALUE

V+ IS THE POSITIVE SAMPLE-AND-HOLD OUTPUT VOLTAGE

V– IS THE NEGATIVE SAMPLE-AND-HOLD OUTPUT VOLTAGE

– RGAIN

RFEF (2(VEXC) – 2(V+ + V–))

2(V+ + V–))

END