Inplementing 60/50 Hz mains rejection (Version 11)
Source : Microchip TC500
https://ww1.microchip.com/downloads/aem ... 21428e.pdf
"The dual slope converter accuracy is unrelated to the integrating resistor and capacitor values as long as they are stable during a measurement cycle.
An inherent benefit is noise immunity. Input noise spikes are integrated (averaged to zero) during the integration periods. Integrating ADCs are immune to the large conversion errors that plague successive approximation converters in high noise environments.
Interference signals with frequencies at integral multiples of the integration period are, theoretically, completely removed, since the average value of a sine wave of frequency (1/T) averaged over a period (T) is zero.
Integrating converters often establish the integration period to reject 50/60 Hz line frequency interference signals. Integration time must be picked as a multiple of the period of the line frequency.
For example, TINT times of 33 ms, 66 ms and 132 ms maximize 60 Hz line rejection.
The threshold noise is the algebraic sum of the integrator and comparator noise. Such errors can be reduced by increased integration times, in the same way that 50/60 Hz noise is rejected."
In my country, Brazil, the mains operates at 60 Hz. In version 10 (initially posted) of the DVM program, I had not yet adjusted the VIN integration time to be a multiple of the 60Hz period.
60Hz period = 1/60 Hz = 16.666 ms
50Hz period = 1/50Hz = 20.0 ms
To increase mains noise rejection, I adjusted the VIN integration time to 16.666 ms. And, incredibly, the measurements became much more stable! Since the integration time was almost doubled, I inserted another 470 nF capacitor in parallel with capacitor C1.
In the case of the offset voltage of the integrating Op amp U3, I replaced the resistor R4 with a 2 K ohm potentiometer. And I adjusted the offset voltage to -15.0 mV.
Obs: If you use an op amp other than OP07, this offset voltage may be different.
Schematics at Github:
https://github.com/Gustavomurta/tinyGo_ ... co_V11.pdf
Source : Microchip TC500
https://ww1.microchip.com/downloads/aem ... 21428e.pdf
"The dual slope converter accuracy is unrelated to the integrating resistor and capacitor values as long as they are stable during a measurement cycle.
An inherent benefit is noise immunity. Input noise spikes are integrated (averaged to zero) during the integration periods. Integrating ADCs are immune to the large conversion errors that plague successive approximation converters in high noise environments.
Interference signals with frequencies at integral multiples of the integration period are, theoretically, completely removed, since the average value of a sine wave of frequency (1/T) averaged over a period (T) is zero.
Integrating converters often establish the integration period to reject 50/60 Hz line frequency interference signals. Integration time must be picked as a multiple of the period of the line frequency.
For example, TINT times of 33 ms, 66 ms and 132 ms maximize 60 Hz line rejection.
The threshold noise is the algebraic sum of the integrator and comparator noise. Such errors can be reduced by increased integration times, in the same way that 50/60 Hz noise is rejected."
In my country, Brazil, the mains operates at 60 Hz. In version 10 (initially posted) of the DVM program, I had not yet adjusted the VIN integration time to be a multiple of the 60Hz period.
60Hz period = 1/60 Hz = 16.666 ms
50Hz period = 1/50Hz = 20.0 ms
To increase mains noise rejection, I adjusted the VIN integration time to 16.666 ms. And, incredibly, the measurements became much more stable! Since the integration time was almost doubled, I inserted another 470 nF capacitor in parallel with capacitor C1.
In the case of the offset voltage of the integrating Op amp U3, I replaced the resistor R4 with a 2 K ohm potentiometer. And I adjusted the offset voltage to -15.0 mV.
Obs: If you use an op amp other than OP07, this offset voltage may be different.
Schematics at Github:
https://github.com/Gustavomurta/tinyGo_ ... co_V11.pdf
Statistics: Posted by Gustavo_Murta — Mon Sep 29, 2025 12:02 am