I forgot to mention but that issue is on ADE7953.

Em 31 de ago de 2017 12:21, "Victor Oliveira" <victor.rsoliveira@gmail.com>

escreveu:

VRMS register (0x21c or 0x31c) is returning a value greater than VPEAK

register (0x226 or 0x326). I've already checked VRMSOS value and it's 0,

meaning that it adds no offset value to VRMS reading, and also I observed V

register (0x218 or 0x318) for instantaneous values of voltage and I see

that the biggest value is equivalent to VPEAK. What is odd to me is that

VRMS value is giving exactly the peak value expected for my application and

VPEAK is giving the exact voltage value in RMS. Here is an example,

VRMS: 5666625

VPEAK: 3980748

Input voltage is 220V RMS at 60Hz. The circuit for voltage channel input

is the following:

>

GNDA is connected to AC_PHASE because I am using a shunt resistor as a

current sensor. I wonder if there is some error in the circuit, but I've

checked another reference circuits and they're pretty similar.

The calculations to convert VRMS and VPEAK values back to real values are

the following:

For VRMS, 5666625/9032007(max value for VRMS) = 0.627, and then 0.627 *

0.5 (full-scale) = 0.313mV. This is the value applied to VN-VP pins and in

my understanding, 0.313mV is the peak value of this signal, not RMS. If I

do the same procedure to VPEAK value (I assume that the maximum value is

the same, 9032007) I get 0.440 * 0.5 = 0.220, and this should be exactly

the RMS value of the input signal.

Does that make any sense? What am I doing wrong?

PS.: I have made the question in the forum, but because I need an urgent

response I also tried via email. I am sorry for duplicates :/

--

Regards,

Victor Rafael S. de Oliveira

Hi Victor,

I see that you are trying to understand your Peak and RMS calculations.

The maximum for the RMS registers are 9032007 but for instantaneous values, in which the VPEAK loop is associated, it is +/- 6500000.

Hope this helps,

Aaron