When using ATT7022EU for energy measurement, the active energy read is different from the theoretical calculation

hello_mcu

When using ATT7022EU for energy measurement, the active energy read is different from the theoretical calculation [Copy link]

Use ATT7022EU for three-phase three-wire energy measurement.

Now the voltage, current and power are all correct. I used a standard source to add a voltage of 57.735 to the three phases and a current of 1A to the two phases A and C. After calibration, the effective values calculated from the voltage and current read from the ATT7022EU are correct; the active power of phase A and the total active power are also correct. The total active power is 173.205W and the phase A is 57.735W.

At this time, I read the active power of phase A and the total active power, but they didn't match. Theoretically, the total active power of the three phases is 173.205W. Because the output voltage and current of the standard source remain unchanged, the power per hour is 173wh, so the total active power per minute should be 173/60=3.7. But when reading the total active power, it takes about 10 minutes to reach 4wh. I don't know what the problem is.

Anyone who has used this chip or has experience please give me some advice.

lugl4313820

Does this require a transformer? How does sampling work?

hello_mcu

lugl4313820 posted on 2023-5-24 08:32 Does this require a mutual inductor? How does sampling work?

Yes. You can directly refer to the application circuit of ATT7022EU to build it.

chunyang

It should be that the coefficient configuration or the related algorithm is not correct.

led2015

Is the time interval for reading total active energy too short?

hello_mcu

chunyang posted on 2023-5-24 16:34 It should be that the coefficient configuration or the related algorithm is not correct.

I forgot to reply. Later I checked and found that what was read was the pulse count, which had to be divided by the HFConst constant.

hello_mcu

led2015 posted on 2023-5-24 19:57 Is the time interval for reading total active energy too short?

It's not because the time is short, but because what is read from the register is the pulse count of the electrical energy, which is not divided by the HFConst constant.