Hi Forum
My application uses an op amp and a comparator. I want to take the output of the op amp, and connect it to a comparator where one input of the comparator is a reference voltage and one input is from the op amp. Since the op amp should be low impedance at the output, and the input of the comparator should be high impedance, I think there should be no issue with signal loss.
My question: Does it make sense to include any resistor value between the op amp and the comparator? If yes, what purpose does the resistor serve, and what value do you recommend?

Thanks
Andy

 

I wouldn't bother but there are other opinions. I believe there was a thread here not long ago that gave reasons for adding the resistor to the input of the comparator.

 

I can remember some variations caused by the input impedances not matching. Input bias current changes with time and temperature, going through the input resistor and causing a difference in switching point, but that's usually millivolts in the worst case. Most of the applications we see here have far worse sources of error than input bias current.

 

Thanks #12, that does jog with my recollection. The issue was input bias current.

 

I would put 1kΩ to 10kΩ on the input of the comparator to limit the input current. I have seen LM311 comparators get blown with no input resistors.
The only negative effect would be slightly slower rise/fall time on the input.

 

I have seen LM311 comparators get blown with no input resistors.

I accede to reality.
"One experimental result is worth a dozen theories."

 

I would put 1kΩ to 10kΩ on the input of the comparator to limit the input current. I have seen LM311 comparators get blown with no input resistors.
....................

I would think that would occur only if the input voltage is above or below the LM311 supply rails (as when its power is off but a signal is applied).

 

I would think that would occur only if the input voltage is above or below the LM311 supply rails (as when its power is off but a signal is applied).

Agreed. I was thinking if the opamp and comparator have the same power supply then you don't need the resistor.

 

My input signal is at 12MHz, so maybe I should worry about the rise and fall times for my application. I think what I'll do is put some 0 ohm resistors there, and if there are problems I can always put more resistance.

 

What comparator are you using?
Is the op amp carrying 12MHz? If so what part is it?

 

the op amp is ADA4940 and it carries a 12MHz signal, which then goes into the AD8611 comparator. do you think these parts are suitable? They seemed fast enough to me..

 

The ADA4940 is designed for differential output.
You might be better is using a conventional high-speed op amp with a single-ended output.
The comparator looks sufficiently fast.
Note that those frequencies require careful layout and decoupling, preferably on a board with a ground plane, to get stable operation and good signal fidelity.