DC blocking capacitor at input of comparator and operational amp

Question

Hi, I simulate with below schematic. LTC6752 is powered by &plusmn; 2.5V. 
 The input signal V3 is 1V AC amplitude with 1V DC bias sinusoidal waveform. 
 We intend to use capacitor to isolate the DC bias and feed only AC waveform to LTC6752 IN+, but the simulation waveform shows that the DC bias is not removed by the cap. 
 Below V(s1) is the original 1V AC amplitude with 1V DC bias sinusoidal waveform. 
 V(s2) is the waveform after capacitor C2. 
 Why DC is not removed? 
 
 And with the same circuit with ADA4522, the DC bias could be removed. (See picture below, right side). 
 
 My simulation file is as below: 
 LTC6752.asc

Achim · Accepted Answer

Hello, 
 your circuit ignores the input bias current, which is required at the input of the opamp/comparator. This is a DC-current, and since your coupling capacitor does not allow DC-currents, the input of the opamp/comparator will drift over time. 
 The simulation looks better for the ADA4522, because its input bias current is much smaller than the input bias current of LTC6752 (in the 100pA range instead of the 1&micro;A range). Due to this huge difference the drift of the input voltage for ADA4522 is 10000 times slower than for the LTC6752 and cannot be recognized on the simulated time scale. But in the real circuit it will appear, both circuits will not work as intended in reality. 
 For a working AC-coupling you have to add some resistor after the coupling capacitor. This resistor will carry the input bias current, and so it will determine the midlevel at the amplifier input. If you take e.g. a 100k resistor after C2 to GND, the midlevel at the input of the LT6752 will typically be in the range of ~100mV (~1&micro;A*100kOhm). 
 best regards 
 Achim)

Partner Forums

Partner Libraries

Products Mentioned

Thread Details

DC blocking capacitor at input of comparator and operational amp