I'm using ADE7912 to measure current(using shunt) and voltage to develop a soft energy meter. The meter is working fine but there's this problem of crosstalk in the voltage channel that I'm facing. Details are as follows:
When I supply current through the calibrator, the voltage on the voltage channel increases(as observed on the debugger). And with increasing the value of current, the voltage further shoots up. However the voltage on current channel doesn’t increase or decrease with increasing or decreasing input voltage through the calibrator. I also verified this by checking the voltage at the voltage channel through a multimeter and the mV value increases with switching on the current supply. I'm attaching some pictures to further illustrate the problem. The voltage values displayed below are measured at the voltage channel of the ADC(VM and V1P) at increasing values of the input current and without changing the input voltage value.
240V input with no current 240V input with 5A current 240V input with 10A current 240V input with 15A current
I would really be thankful if you could provide your inputs on this and help me out by suggesting ways to remove this crosstalk issue.
Can you supply me with your schematic and pictures of your sensor connection.
Sure Dave. Here you go.
connection of ADE7912 with the uC
MCU Schematic where ADE is connected
phase, neutral and shunt signals S+,S- and Ref are coming from another board connected through connector JP3 in the above. The shunt schematic is attached below as well
Hope this much information will suffice. In case you need anything more than this please let me know.
Thanks! Can you please provide more information on how the VDD, GroundF supply is generated? Also how big is your shunt resistor?
VDD and Ground are generated using an offline SMPS which gives an output of 5V and 3A. This 5V is further connected to an LDO (LDK220) which generates 3.3V(VDD in this case). SMPS is on one PCB and its 5V output is taken to another PCB through a header where 5V is converted to 3.3(VDD) using LDK220. Attached herewith is the connection of LDK.
The SHUNT value used is 1 milli-ohm
Please let me know if more information is required. Awaiting your response eagerly.
It looks like you have a ground loop where the VM pin is connected to PHASE_IN and AGND, through the anti-alias filter. AGND is connected to PHASE_OUT through the ferrite bead to the shunt.
Try to improve this so you only have one connection to AGND, coming in from the shunt.
You can also try removing the shunt from the connector and shortening the connection from the shunt to the inputs of the ADE IC as much as possible to see if this improves the performance. (Having as short as possible a connection here is really important.)
Try using an isolated benchtop power supply to supply 5V directly instead of through the SMPS to rule this out if you are still having issues.
Please report back and let us know how you make out.