Please, explain the capacitor in the DC BIAS and overcurrent protection

Studying the DC BIAS diagrams in the site I see that they use a resistor divisor to get the DC BIAS value. Trivial enough. But I don’t understand why are you using a capacitor to ground. Several explanations say that this is used to provide a ground for AC component but I don’t understand the explanation.

Yes, I understand that the capacitor is a open circuit for DC and a short circuit for AC (ideally) but I just don’t understand WHY is that needed at all.

I even see this schema in the OpAmp version, where it makes even less sense to me.

I have the feeling that I am missing something important.

Please, could you state the detailed reason for this?. I don’t understand the “it provides a ground for AC” explanation. I get the physical meaning but I don’t understand the reason why that is needed.

In fact, the interaction between the voltage divider and the inductor in the CT looks complicated and I don’t see any analysis about it. Thinking about it, this could be the exact reason for the capacitor, isolating the divider from the inductor, providing a low resistence path to avoid disturb the voltage divisor. There is a non trivial interaction, though.

Anyway, isolating the voltage divider from the CT seems a pretty good idea (via an OpAmp).

I don’t see either any overcurrent protection. That is, what happens if a 30A rated CT is measuring a 35A current. Apparently you would destroy the ADC pin in the chip and blow the capacitor if it has polarity (electrolytic). Could be that the magnetic core of the CT gets saturated and you are counting on that to protect you?.

I like to understand what I am doing instead of just cloning designs without any real understanding. I am engineer, although not in electronics :-). Have been 25 years since I did a SPICE simulation…


Firstly, the CT doesn’t look much like an inductor, it behaves much more like a current source. The burden resistor turns it into a voltage source, which has a fairly low source impedance.

Second, the ADC isn’t simply an ADC, it has a multiplexer in front of it, which switches between the 6 available input pins. The multiplexer samples the voltage and stores it on the sample & hold capacitor, to be converted by the ADC.

In order to charge that S&H capacitor sufficiently quickly to ensure an accurate reading, the Atmel data sheet tells us that the driving impedance needs to be less than 10 kΩ.

The combination of CT and burden resistor easily satisfies that requirement, but when you add in the Thévenin equivalent of the bias resistors (1.35 V with a source impedance of 235 kΩ for the emonTx), that limit is exceeded by a large margin. Adding the bypass capacitor provides a source to supply the charge that the S&H capacitor needs.

In terms of overcurrent protection, there are protection diodes inside the ADC inputs, which are in turn protected by a 1 kΩ series resistor and parallel transient voltage suppressor diodes (in production versions of the emonTx and emonPi) and yes, by the CT itself saturating. The impedance of the bias network also helps to limit the current in the protection diodes.

If you’re copying a design, you should probably copy the production version. The designs in Building Blocks work, but they are really only intended to explain the principles.