JAJSE50 November 2017 TLA2021 , TLA2022 , TLA2024
PRODUCTION DATA.
The TLA202x use a switched-capacitor input stage where capacitors are continuously charged and then discharged to measure the voltage between AINP and AINN. The frequency at which the input signal is sampled is referred to as the sampling frequency or the modulator frequency (fMOD). The TLA202x have a 1-MHz internal oscillator that is further divided by a factor of 4 to generate fMOD at 250 kHz. The capacitors used in this input stage are small, and to external circuitry, the average loading appears resistive. Figure 9 shows this structure. The capacitor values set the resistance and switching rate. Figure 10 shows the timing for the switches in Figure 9. During the sampling phase, switches S1 are closed. This event charges CA1 to VAINP, CA2 to VAINN, and CB to (VAINP – VAINN). During the discharge phase, S1 is first opened and then S2 is closed. CA1 and CA2 then discharge to approximately 0.7 V and CB discharges to 0 V. This charging draws a very small transient current from the source driving the TLA202x analog inputs. The average value of this current can be used to calculate the effective impedance (Zeff), where Zeff = VIN / IAVERAGE.
The common-mode input impedance is measured by applying a common-mode signal to the shorted AINP and AINN inputs and measuring the average current consumed by each pin. The common-mode input impedance changes depending on the full-scale range, but is approximately 6 MΩ for the default full-scale range. In Figure 9, the common-mode input impedance is ZCM.
The differential input impedance is measured by applying a differential signal to the AINP and AINN inputs where one input is held at 0.7 V. The current that flows through the pin connected to 0.7 V is the differential current and scales with the full-scale range. In Figure 9, the differential input impedance is ZDIFF.
Consider the typical value of the input impedance. Unless the input source has a low impedance, the TLA202x input impedance may affect the measurement accuracy. For sources with high-output impedance, buffering may be necessary. Active buffers introduce noise, offset, and gain errors. Consider all of these factors in high-accuracy applications.
The clock oscillator frequency drifts slightly with temperature; therefore, the input impedances also drift. For most applications, this input impedance drift is negligible and can be ignored.