hadi.esfandi
Newbie level 1
Hello friends
I am working on designing analog front-end of a circuit include high resolution ADC. In order to reduce the cost, we intend to use adc low price (MCP3918 Microchip). for the first step I should decrease the amplitude of the signal to be in range of the adc and I do this work by resistor deviding. Is it appropriate approach? I think it increases the noise level of the signal because of resistor noise(4ktr) and decrease the SNR of the signal.
But the main question is about spikes which are appeared in the signal. If I set OSR of the adc to 32 i see many spikes( about 8 in one minutes) but If I set to 4096 the condition will be better and the number of the spikes will be getting low(about 1 in 10 minutes).however when I see digital signals in matlab the spikes are just one sample and this issue reduce the possibility of analog spikes before adc and I think these spikes are formed in adc. The attached file is the schematic of the adc section of our circuit and spikes which appear in digital signal.
Is there anybody who know about origin of these spikes and how to eliminate them without digital approaches?
I am working on designing analog front-end of a circuit include high resolution ADC. In order to reduce the cost, we intend to use adc low price (MCP3918 Microchip). for the first step I should decrease the amplitude of the signal to be in range of the adc and I do this work by resistor deviding. Is it appropriate approach? I think it increases the noise level of the signal because of resistor noise(4ktr) and decrease the SNR of the signal.
But the main question is about spikes which are appeared in the signal. If I set OSR of the adc to 32 i see many spikes( about 8 in one minutes) but If I set to 4096 the condition will be better and the number of the spikes will be getting low(about 1 in 10 minutes).however when I see digital signals in matlab the spikes are just one sample and this issue reduce the possibility of analog spikes before adc and I think these spikes are formed in adc. The attached file is the schematic of the adc section of our circuit and spikes which appear in digital signal.
Is there anybody who know about origin of these spikes and how to eliminate them without digital approaches?