Hello

I am using a board with current transformer and AC/AC voltage transformer with ADE7753. While everything works, calibration gives me some major problems. I am using an accurate source, but I would also try to calibrate it with a reference meter as soon as I got most of the problems solved. Here is what I am struggling with:

- VRMS register returns a value of about 26000 decimal when the V2P/V1N inputs are internally shorted and there is no signal. I tried with various combinations of CH2OS and VRMSOS, and so far the most effective is CH2OS = 0x24 and VRMSOS = -2048. This gives a value of around 1700 dec in VRMS for no signal condition. I want to make VRMS equal to zero when there is no signal; also ano other number loaded in CH2OS makes the VRMS value go higher and higher. On the other hand, I am able to make the IRMS register stay at zero for no current condition.

- I tried calibrating the meter several times using various loads with PF = 1. The problem after calibrating is that the LAENERGY register contains a value which is slightly higher than the contents of LVAENERGY register (both are scaled using the Wh/LSB and VAh/LSB constants). For example, for a 1 kW load with PF = 1, the values are 1040 W and 1035 VA. When I multiply and scale the current and voltage signals for the same load, I get 1080 VA.

Every time I perform a calibration I end up with more or less the same values for CFDEN and WGAIN. I always have a value of 0 for VAGAIN; is this a strange behavior? Here are the results of my last calibration:

CFDEN = 1916 dec

WGAIN = -2007

VAGAIN = 0

APOS = 23000

Wh/LSB = 0.00016301512780386018

VAh/LSB = 0.00009790546761651741

The PCB configuration is taken from the datasheet and everything else works as expected. I use the routine described at page 47 of the datasheet for calibrating VAGAIN and WGAIN simultaneously. Now I am stuck with this for days and can't continue working on my project. Any help will be greatly appreciated!

Regards

Hi TeddyB,

I apologize for the delay in the response. First off, instead of shorting the inputs internally with the bit, just apply the lowest level AC voltage you expect on the line and perform an offset calibration with only the VRMSOS register, not using the CH2OS register. I would never expect you to get exactly 0 in the VRMS reading with 0 volts at the inputs because there will always be some amount of noise which will show up as a small amount in the RMS register.

With regards to the scaling of Active vs. Apparent, the two registers don't have the same scaling factor. This can be seen in Table 10 of the Data Sheet on Page 38. Therefore, if you want them to display the same value for PF = 1 then the gains will have to be adjusted accordingly to give the same Wh/LSB or VAh/LSB constant for both.

Apparent = Active * 0.848

I would recommend doing the calibration in a slightly different order than described in the data sheet. First, start with a phase calibration if you are performing one. Then perform the gain and offset calibration on the RMS registers. Finally perform the gain and offset calibration on the power.

As mentioned above, perform the offset calibration with the smallest expected voltage signal for VRMS, the smallest expected current for IRMS and Energy.

Regards,

Dlath