ADE9000 and Voltage output CTs

 dave.smith  I think I found a post that clarifies based on your 25y experience,  RE: Regarding Current channel sensing through CT in ADE7953 

Is this applicable to all AFEs we are using: ADE9000, ADE9078, ADE9430 and ade 7953?

What issues are introduced using the setup1 (single-sided with one leg to GND)?

Thank you.

Domenico

Ok there are a few ways to deal with this as you have found in the link.

The goal is to not float the inputs and tie something to AGND. In the diagrams below the first 2 are relevant to you. Consider R1 is inside the CT to give you a current in voltage out CT. You can tie one leg to gnd a shown in the first figure or use 2 higher value resistors tied to each leg to gnd shown in the  second figure. the 2 higher value resistors are in parallel with the internal ct burden so you need to make sure they don't change much with temp and change the transfer function of the CTover temp. Try both see what works best for you. Create your schematic using figure 2 then it is easy to remove a resistor and change teh other to 0 ohms to test figure 1. This will work for all of our parts that use CTs as the current sensor.

If the inputs are not ties to gnd some how the common mode will be an issue and can cause distortion if it gets to big. 

When you single end the current measurement as in the top figure any signal that couples on the trace/lead from the CTconnection to the adc can be measure by the  ADE. Twist your CT wires tight to improve noise rejection before the CT wire is grounded to agnd. 

the middle figure is better as signal are more differential can common mode noise can be cancelled by the ADC inputs.

I prefer the  burdens on board instead of voltage output CTs since the current loop from the ct to the burden is difficult to disturb and with differential burdens like the bottom figure noise performance is really good 

Dave

Thank you Dave for the exhaustive explanation.

Thank you Dave for the exhaustive explanation.