# Does input impedance of an isolation amplifier or a transformer have impact on common mode to differential mode conversion?

I have a question where I can only explain it clearly by an example. It is about conversion of common-mode interference to differential mode due to overall system imbalance.

I will use below two models where a common mode voltage is applied through a cable to amplifiers in both cases. In Figure 1 differential amplifier is the receiver; and in Figure 2 an isolation amplifier or any system which has isolation like a transformer is the receiver. Let's first look at Figure 1. Rs1, Rs2 can be source output resistances and C1, C2 can be the cable's capacitance to the ground. Rin1 and Rin2 represents the input impedance of the differential amplifier. Here it is apparent that if we want to estimate how much Vcm is converted to differential mode then we need to know Rs1, Rs2, C1, C2 and also Rin1 and Rin2. We can also see that even Rin1=Rin2, still value of input impedance have an impact on the value of common mode to differential mode conversion.

Now let's examine the case for Figure 2 where the isolation amplifier is used ans same Vcm is present. For simplicity I tried to model as it has input impedance Rin and parasitic isolation capacitance as Cp1 and Cp2 as shown. In this case it is to me apparent that Rs1, Rs2, C1 and C2 again determines the conversion from CM to DM. But it is not very obvious whether the value of Rin has impact on the value of the common mode to differential mode conversion.

Regarding Figure 2 my questions are:

It seems the common mode currents will not flow through Rin but they will flow through Cp1 and Cp2. Is that correct? If so does that mean the input impedance value of the isolation amplifier has no effect on CM to DM conversion and act almost as an open circuit to the common mode voltages?

## 1 Answer

If Rin is low in value, the voltages at the input terminals of the isolation amplifier will be "dragged" closer to each other compared to the scenario when Rin is high in value. This means that Rin's value does affect the conversion of common mode to differential mode signals. If you take it to extremes, if Rin is zero there can be no differential voltage.

• When there is isolation barrier I'm not able to estimate the CM to DM conversion because I dont know what Cp1 and Cp2. For example from this data sheet:dataforth.com/catalog/pdf/8b30_31.pdf I treat the amplifier input as open to CM voltages to estimate the CM to DM conversion but theoretically I cannot prove it to my self. That's why I asked the question whether my thinking is correct. – user1245 Sep 9 '19 at 13:45
• Only differential currents flow through Rin but that differential current can be due to imbalances in the other components converting common mode signals to differential signals. Rin DOES affect that conversion. – Andy aka Sep 9 '19 at 14:14
• Thanks for the answers. Is my Figure 2 correct enough to use to estimate the CM to DM conversion? I now simulated and Rin really have effect especially if it is low more effect. They give it as 650k but Cp1 and Cp2 I dont have. – user1245 Sep 9 '19 at 14:34
• Yes, the figure 2 is a good start. At higher frequencies (over 100 kHz) it's probably a good idea to model cable series inductance as this will have an effect on things and make high frequencies more attenuated compared to low frequencies. – Andy aka Sep 10 '19 at 7:00