Hello friends!
I am using an OPA671 OPAMP in unity gain mode (as a buffer). Problem is that, sometimes during
the operation it suddenly starts oscillating. I dont know why but as per the datasheet I put a
parallel combination of resistor and capacitor at the output of it, but still not able to
retify the problem. I am attaching the schematic. Plz give some comments on it.
Thanx.

Originally I am using 2 OPA671 as a part of a peak detector circuit wherein output of 2nd OPA (Buffer)
is beeing fed into first OPA.I am also attaching the schematic of interconnection of two OPAMPS.

Can u suggest something which can prevent oscillations, and improve the performance of the circuit..

 

Why do you fed OPA2 back to OPA1?

Where did you find the peak detector circuit?

It does not look right to me.

 

Is there a parallel RC network from the cathode of D4 to GND that is missing from the schematic?

 

Are you building this on a zip-strip type breadboard?

 

sorry dear..

yes there is an RC network from D4 to ground, but that is series not parallel. I am attaching the new schematic.

At MIK3..
Second OPA is a buffer. First one is peak detector.

 

This circuit is OK.
Some time it works pretty fine. Only if I switch on and off the power supply it goes into oscillations.

LM319 generates the detect signal. I am attaching the new schematic.

 

You have no provisions to discharge the input of U2 pin 3.

If a peak is detected and held on that input, then power is turned off and back on, you may have oscillations.

You need to figure out a way to discharge that input when turning off the power.

 

thats just the analog part.. I am having a current source to discharge the capacitor after each peak is detected. I just posted the analog part that oscillations are due to opamps only.

 

Well then, you're on your own to figure it out.

Your schematics have been difficult to read, you don't post the entire circuit, and we have NO clue what your board layout is - or even if it IS on a PCB vs a breadboard.

The only item that make me suspect that it is on a PCB is the BAS81's, which would not be usable on a breadboard.

 

To find out the answer I went through a lot of browsing and came across this

http://www.allbusiness.com/media-telecommunications/data-transmission-bandwidth/6240674-1.html

They are suggesting using two opamps with different BW. can u suggest a pair which fulfill the condition. Or any other idea...

 

I put 1 capacitor (220p pf) and a resistor (47k) i tionshe feedback loop of second opamp (buffer) and oscillations are gone. This idea was given in datasheet.

nyway thanking u guys for ur suggestions

 

The second opamp must have a bandwidth that is much higher (low phase shift) than the first opamp.

If you made the circuit on a breadboard then its high capacitance between strips of contacts and wires all over the place cause the circuit to oscillate.