Many system designers use ∑ – Δ ADC and RTD (resistance temperature detector) for temperature measurement, but it is difficult to achieve the high performance specified in the ADC data book. For example, some designers may only get 12 to 13 noiseless bits from 16 to 18 Bit ADCs. The front-end technology introduced in this paper can enable designers to obtain more than 16 noiseless bits in their system design.

The use of RTD in ratio measurement has certain advantages because it can eliminate the error sources such as precision and drift of excitation current source. The following is a typical circuit for 4-wire RTD ratio measurement. The advantage of the 4-wire configuration is that it eliminates errors caused by pin resistance.

Excellent analysis of different analog front end designs

Figure 1. 4-wire RTD ratio measurement circuit.

We can derive the following two formulas from the above circuits:

Excellent analysis of different analog front end designs

When the ADC operates in bipolar differential mode, the general expression for calculating the RTD resistance (rrtd) is as follows:

Excellent analysis of different analog front end designs

Among them:

Codertd is ADC code.

CodeADC_ Fullscale is ADC full scale code.

Theoretically, the measured resistance of RTD is only related to the accuracy and drift of the reference resistance. In general, rref is an accurate low drift resistor with an accuracy of 0.1%.

When engineers design products with such circuits, they add resistors and capacitors in front of analog input and external voltage reference source pins to obtain low-pass filtering and overvoltage protection as shown in Figure 2. In this article, we will show the factors that should be considered when selecting the appropriate resistor and capacitor for better noise performance.

Excellent analysis of different analog front end designs

Figure 2. Typical 4-wire RTD ratio measurement circuit.

As can be seen from Figure 2, R1, R2, C1, C2, and C3 are used as first-order low-pass RC filters that provide attenuation as differential and common mode voltage signals. The values of R1 and C2 should be the same. Similarly, R3, R4, C4, C5, and C6 are used as low-pass filters for the reference path.

Common mode low pass RC filter

Figure 3 shows the equivalent circuit of the common mode low-pass filter.

Excellent analysis of different analog front end designs

Figure 3. Common mode low pass filter.

Because the common mode voltage at point a is equal to the voltage at point B, no current flows through C3. Therefore, the common mode cutoff frequency can be expressed as:

Excellent analysis of different analog front end designs

Differential mode low pass RC filter

In order to better understand the cut-off frequency of low-pass RC filter for differential signals, the C3 capacitance in Fig. 4 can be regarded as two independent capacitors in Fig. 5: Ca and CB.

Excellent analysis of different analog front end designs

Figure 4. Differential mode low pass filter.

Excellent analysis of different analog front end designs

Figure 5. Equivalent circuit of differential mode low pass filter.

In Figure 5, the cut-off frequency of differential mode is as follows:

Excellent analysis of different analog front end designs

In general, the value of C3 is 10 times that of CM. This is to reduce the impact of C1 and C2 inconsistencies. For example, as shown in Figure 6, when analog front-end design is used in ADI circuit notebook cn-0381, the cut-off frequency of differential signal is about 800 Hz, and that of common mode signal is about 16 kHz.

Excellent analysis of different analog front end designs

Figure 6. Analog input configuration for RTD measurement using ad7124.

Resistance and capacitance considerations

In addition to being a part of the low-pass filter, R1 and R2 can also provide over-voltage protection. If a 3K Ω resistor is used in front of the ad7124-4 ain pin in Figure 6, it can protect up to 30 V wiring errors. It is not recommended to use larger resistance in front of AIN pin for two reasons. First, they will produce more thermal noise. Second, the AIN pin has an input current that will flow through these resistors and introduce errors. The magnitude of these input currents is not constant, and mismatched input currents will produce noise, and the noise will increase with the increase of resistance.

Resistance and capacitance values are critical to determine the performance of the final circuit. The designer needs to understand the site requirements and calculate the resistance and capacitance values according to the above formula. For ADI ∑ – Δ ADC with integrated excitation current source and precise analog microcontroller, it is recommended to use the same resistance and capacitance values before ain and voltage reference pins. This design ensures that the analog input voltage is always proportional to the reference voltage, and any error of the analog input voltage caused by the temperature drift and noise of the excitation current can be compensated by the change of the reference voltage.

Noise performance of aducm360 measured by ratio measurement method

Aducm360 is a fully integrated 3.9 ksps, 24 bit data acquisition system. It integrates dual channel high-performance multi-channel ∑ – Δ ADC, 32-bit arm? Cortex? – m3 processor and flash / EE memory on a single chip. At the same time, programmable gain instrument amplifier, precise bandgap voltage reference, programmable exciting current source, flexible multiplexer and many other features are integrated. It can be directly connected with resistance temperature sensor.

When using aducm360 for RTD measurement, the ref – pin is usually grounded, so there is no current flowing through R4 and C5 in Figure 2, which can be removed. C4 and C6 are connected in parallel. Since C4 is much smaller than C6, it can be ignored. Finally, a simple analog front-end circuit can be obtained, as shown in Figure 7.

Excellent analysis of different analog front end designs

Figure 7. Aducm360 analog front end circuit for RTD measurement.

Table 1 lists the noise levels with matched and unmatched filters in front of the analog and reference input paths. Use 100 Ω precision resistor instead of rrtd to measure noise voltage on ADC input pin. The value of rref is 5.62 K Ω.

Table 1. Noise Test Results

Excellent analysis of different analog front end designs

From table 1, we can see that when using the matched analog front-end circuit with the same values of R1 and R2 as R3, the noise is reduced by about 0.1 μ V to 0.3 μ V compared with the unmatched circuit, which means that the number of noise free bits of ADC increases by about 0.25 to 16.2 bits, and the gain of ADC PGA is 16.

conclusion

According to the considerations described in this paper, using matched RC filter circuit and selecting appropriate resistance and capacitance values according to site requirements, RTD in ratio measurement application can obtain the best results.

Leave a Reply

Your email address will not be published. Required fields are marked *