When we choose the sample frequency 100SPS - thee will not be any 50Hz noises (but it may be any very-low frequency fluctuations - because difference between, said, 50,1 input frequency and sampling frequency, as a result we will have periodic signal with period 10Sec and shape like rectified sine wave ).
Part 1 was well delivered but Part 2 completely lost me here - are Dual-Slope and Charge Balancing techniques for circuit implementation (where we have to construct the circuit) or do they exist as a package? You can purchase an SAR ADC chip, but I don't see any vendor selling ADC categorized as Dual-Slope ADC or Charge-Balancing ADC.
Alle these implementations can be found in packages. You are right that you can purchase an SAR ADC by searching for it, but "Dual-Slope" and "Charge-Balancing" are just two techniques used in integrating ADCs. Manufacturers only sometimes specify which technique is used in their ADCs. For example this ADC by Analog Devices, which uses charge balancing: www.analog.com/media/en/technical-documentation/data-sheets/AD7710.pdf ...or this one by Micorchip, which uses a dual-slope approach: ww1.microchip.com/downloads/en/DeviceDoc/21428e.pdf
Regarding time averaging, let's assume that input signal has AWGN & we want to calculate the noise power in the system, which element is going to determine the cutoff frequency for noise PSD ? LPF before the ADC ? Or averaging as shown in 3:34 ? Thanks
This is a quite tricky question, since there are many more factors involved when it comes to integrating ADCs. I would assume that you would want your low pass filter in front of the ADC to determine the cutoff frequency. If you want your ADC to surpress e.g. humming in the lower freqeuncy range, the averaging-process will also play a role. But it will also depend on things like the order of the ADC and - when it comes to delta-sigma-ADCs - the oversampling frequency of the signal and the type of filter that is used for windowing the time-domain-signal (see part 3 of the series for explanation). So I'm sorry to say that I have no clear answer to this question.
When we choose the sample frequency 100SPS - thee will not be any 50Hz noises (but it may be any very-low frequency fluctuations - because difference between, said, 50,1 input frequency and sampling frequency, as a result we will have periodic signal with period 10Sec and shape like rectified sine wave ).
Nicely explained 👍
Part 1 was well delivered but Part 2 completely lost me here - are Dual-Slope and Charge Balancing techniques for circuit implementation (where we have to construct the circuit) or do they exist as a package? You can purchase an SAR ADC chip, but I don't see any vendor selling ADC categorized as Dual-Slope ADC or Charge-Balancing ADC.
Alle these implementations can be found in packages. You are right that you can purchase an SAR ADC by searching for it, but "Dual-Slope" and "Charge-Balancing" are just two techniques used in integrating ADCs. Manufacturers only sometimes specify which technique is used in their ADCs. For example this ADC by Analog Devices, which uses charge balancing:
www.analog.com/media/en/technical-documentation/data-sheets/AD7710.pdf
...or this one by Micorchip, which uses a dual-slope approach:
ww1.microchip.com/downloads/en/DeviceDoc/21428e.pdf
Regarding time averaging, let's assume that input signal has AWGN & we want to calculate the noise power in the system, which element is going to determine the cutoff frequency for noise PSD ? LPF before the ADC ? Or averaging as shown in 3:34 ?
Thanks
This is a quite tricky question, since there are many more factors involved when it comes to integrating ADCs. I would assume that you would want your low pass filter in front of the ADC to determine the cutoff frequency. If you want your ADC to surpress e.g. humming in the lower freqeuncy range, the averaging-process will also play a role. But it will also depend on things like the order of the ADC and - when it comes to delta-sigma-ADCs - the oversampling frequency of the signal and the type of filter that is used for windowing the time-domain-signal (see part 3 of the series for explanation). So I'm sorry to say that I have no clear answer to this question.