Hallo, i need a high pass filter with a cutoff of 0.5Hz to cancell out a DC component from a signal.(signal = 1.5V DC +5mV AC ).
I'm thinking that i simple 1 pole RC filter would serve for this purpose, but how to chose between many R and C value pairs?
i mean i can have 4Mohm resistor + 0.1uF capacitor, or 1uF capacitor and 400KOhm resistor and so on.. How to choose the best one, or it doesn't matter as long as capacitor is not too big in size for e.g?

 

Generally, any capacitor in series would block DC.

 

yes, but there is an op-amp in the following stage, so i think using only capacitor wouldn't work. I found an example

 

The choice of the resistor is determined by the input bias current of the OpAmp that follows. Say the input Bias current spec is 10nA. If you used a 1megΩ resistor, then the voltage drop across this resistor is I*R = 1e-8*1e6 = 1e-2 = 10mV. This means that you will have an offset at the output of the OpAmp of 10mV * Gain of the OpAmp. If this is a problem, then make the resistor smaller and try the calculation again.

 

The choice of the resistor is determined by the input bias current of the OpAmp that follows. Say the input Bias current spec is 10nA. If you used a 1megΩ resistor, then the voltage drop across this resistor is I*R = 1e-8*1e6 = 1e-2 = 10mV. This means that you will have an offset at the output of the OpAmp of 10mV * Gain of the OpAmp. If this is a problem, then make the resistor smaller and try the calculation again.

Thanks, you really helped me. Offset is a big problem. but how do the op amp's input bias current flow thru the filter resistor (for e.g in that example schematic)?

 

The opamp isn't ideal and draws current. This is how it flows through the resistor to the opamp terminals.

 

... Offset is a big problem. but how do the op amp's input bias current flow thru the filter resistor (for e.g in that example schematic)?

The bias current CANNOT flow through the series capacitor; ero it HAS TO flow through the resistor between the OpAmp input and ground (there is not other path where it can flow). Some OpAmps, the bias current flows INTO the input pin; some, the current flows the other way. Only the OpAmp's Data Sheet knows for sure.

 

A common mistake in such designs I believe is omitting a DC return path resistor.

 

A common mistake in such designs I believe is omitting a DC return path resistor.

yes, i've read that in such case bias current charges the capacitor till ir saturates OpAmp

 

The bias current CANNOT flow through the series capacitor; ero it HAS TO flow through the resistor between the OpAmp input and ground (there is not other path where it can flow). Some OpAmps, the bias current flows INTO the input pin; some, the current flows the other way. Only the OpAmp's Data Sheet knows for sure.

The input bias current flows FROM OpApm's input to ground. Did i undestood corectly?

 

The input bias current flows FROM OpApm's input to ground. Did i undestood corectly?

Depends on the specific opamp. Both types exist.

 

I would like to ask another question, i think it's somehow related to this topic so to not start another thread i ask here.
I need some low power OpAmps for a battery powered device(two 3V CR2032 coin batteries). I found OpAmps that is said to draw ~10 uA of current. To my knowledge it's a current with no load attached.So to make my OpAmps really low power(current per amplifier <0.1mA), i need to choose very large feedback loop resistors?(I need a gain of +200) Or does supply the current depends on something else?

 

You're almost always better off starting a new topic, if it is not exactly the same as the 1st question posed in the thread. Otherwise, the thread can get confused very quickly.

Using large values of resistance can help to reduce power consumption, but the higher the resistance, the more susceptible to noise the circuit will be - and more noise will be created by the resistors themselves. As an analogy, think of how noisy a garden hose nozzle is when you have it set to spray a jet, vs a garden hose with no nozzle.

Metal film resistors are less noisy than carbon resistors.