Problem with the VRMS and IRMS working because of ADE7753 grounding

[markdem]

Hi All, I have finally got my ADE7753 talking to my PIC, I can read all the non-zero registers, get the die revision and read the period time.

I cant, however, get the VRMS and IRMS working. I can read the register OK, but i get unstable numbers, ranging from 20000 to 60000, even when I short the inputs. I think that this may be a hardware problem. I have looked at AN564 from analog, and there is something that I am not to sure about.
In the diagram for the power supply, it shows the analog ground connected to the 240VAC neutral line, there for connecting the ground from the 7805 to it. Is this correct?

Currently I have a simple 7805 power supply for the PIC, ADE7753 and my LCD, and then I have a resistor divider to reduce the 240VAC down to 470mV for the ADE7753.
The regulator and resistor divider and not connected in any way. Is this how it should be, or am I getting something very wrong.

Thanks

Mark

 

[FvM]

ade7753 neutral analog ground

In the ADE7753 application circuits given in the datasheets as well as in the AN564, the AGND to neutral or L2 connection is required. I wonder how you presented the voltage measurement to V2P/V2N without the ground connection. There may be the option to use a differential voltage divider and no ground connection. This would allow to keep the circuit ground isolated. But unfortunately, no common made rejection is specified for the differential input.

 

[markdem]

ade7753 & vrms

Thanks FvM. Is this what I need to do with the power supply and resistor divider? I have left out the caps, rectifier and MOV around the power supply, just to make it simple.

Thanks, Mark

 

[markdem]

ade7753 voltage measurement

OK, I have now done what I suggested above, but the voltage reading is still very unstable. It will go between 242V to 250V in about 6 seconds. It looks like it always goes under the real reading, eg sometimes when the multimeter read 247V, the ADE output is 243V. I am measuring the line voltage, and it is very unstable, but the multimeter does not move as much so fast.
Is this a problem, or is it just because the ADE7753 is very accurate\not accurate? I guess I could average it out in the PIC, but I was hoping for a stable reading from the ADE.

Thanks again, Mark

 

[FvM]

vrms divider ade7753

I think, you forgot to draw some rectifier diodes, otherwise the transformer is hopefully shortcircuit proof... Apart from that, I don't see a particular reason for an unstable voltage measurement. The chip should have an accuracy better than 1% in any case. So either a hardware problem or incorrect programming should be expected.