With such high value resistors on the opamp inputs the circuit is sensitive to interference. Your body acts as an antenna for picking up interference which can be coupled capacitively to the opamp inputs.
Is the circuit inside a screened enclosure? If not, it should be.
With such high value resistors on the opamp inputs the circuit is sensitive to interference. Your body acts as an antenna for picking up interference which can be coupled capacitively to the opamp inputs.
Is the circuit inside a screened enclosure? If not, it should be.
the comparator is !basically! an open loop op amp make your schematic first with these (inexpensive ones) see how it acts
the schematic to work fast takes biasing of near minimum allowed resistance values also minimum not to "distort" your input (by dc bias e.g. if your comparator input biasing/compensation network is 100 x MΩ range but your source is 1 x GΩ and the ΔVin is small it might have no NRG to drive the detector or to drive it properly)
if your input is high impedance you likely want to buffer(filter/alter) it for the same reason about impedance converters https://en.wikipedia.org/wiki/Buffer_amplifier to match your preferred biasing/compensation for comparator -- the drawback is you introduce an additional time delay and error-function to the input chain
◄ this is when you use op amp for comparator or to get the hint what's going on in biasing/feed-back network (as the difference of referencing your multimeter's COM to variety of potentials avail around your test circuit or modifying it's probes with series resistors or j-Fet input translator/buffer)
notice the "Problem 6.3 - Follower" + "Problem 6.7 - Even More Virtues of Feedback - I" & "Problem 6.8 - Even More Virtues of Feedback - II" in https://instrumentationlab.berkeley.edu/Lab6