I got a few minutes to work on the ADE7763 energy monitoring chip. Now that I am confident in the SPI interface, I started to look at the functioning of the chip itself. I started with the test of having the ACD inputs for voltage and current shorted to ground as if there was no voltage or current being measured. As mentioned in a previous post, I was getting non-zero reading from the Irms and Vrms registers.
A reply from a query to Analog Devices let me know that this was expected. That with very low or zero input signals, the output will not be zero.
So the next test was to see if those “zero” values would change if the common mode reference was not referenced to ground but to other voltages. The answer is that they do but not by much.
Next step I took was to apply a variable DC voltage to the plus input of one of the DACs to see if I could vary the corresponding RMS register. The minus input was referenced to ground. I did get some variations but with the full scale voltage input, I was not getting the max 24 bit output in the register. So I looked at the internal processing chain to see what register setting I may have overlooked when it dawned on me what the problem is.
This chip will not measure DC voltages. With all the low and high pass filters, zero crossing detectors, and what not, this chip only measures AC voltages. So the next step was to rig up a 60Hz variable voltage test signal to the ADC input. And now the appropriate RMS register changes in proportion to AC input.
Next step will be to check the linearity of the ADC inputs to the register values. After that it will be to connect up a real CT to the current register.
At the moment, this looks very promising.