• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 49 subscribers
  • Views 572 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

OPA659: High Speed Peak Detector

OPA659: OPA 659 Overshooting or Ringing Problem

Behtan Umut Mercan
Prodigy 90 points
Part Number: OPA659
Other Parts Discussed in Thread: TINA-TI

Hello,

I have designed a peak detector circuit with OPA 659. My purpose was to measure the peak of the fast (about 300 kHz) signal. Therefore I connected the output of the opamp to microcontroller ADC. In order to get more accurate results, I averaged 20-50 values of ADC samples. 

However the results for fast signal were not true. For example, when I intended to measure 1000 V I got 1400 V from the microcontroller. I guess my opamp output reaches even higher than 1400 V because of averaging. (It worked fine with dc values.)

I think the reason for that is my opamp rings or overshoot the values.

What could be the reasons for this situation? 

What should I do to solve it?

I attached my schematic below. 

Sincerely,

Behtan Umut Mercan

Peakdetector.TSC

  • 0
    Guru 140200 points

    Hi Behtan,

    your simulation does not show a 300kHz signal but only 42kHz. Can you tell us your actual input signal?

    Why have you added R1?

    You should think about using a frequency compensated input voltage divider. Otherwise R6 and C3 will work as a low pass filter with 850kHz.

    What ADC are you using?

    Kai

  • 0 in reply to kai klaas69
    Prodigy 90 points

    Hi Kai,

    Thank you for your answer.

    My signal is a surge signal with a rise time of 1 microsecond. Considering 300 kHz signal has rise time less than 1 microsecond I wrote like that. I attached my waveform below. It will repeat every 10 seconds.

    I have added R1 as an isolation resistor to opamp 2. And R7 and C2 as low pass filter.

    I have mentioned you about my resistor divider in the below thread. Since I am using 1206 resistors with low stray capacitance I thought my signal would not get distorted. 

    (https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/1034591/opa659-high-speed-peak-detector)  

    For the analog-digital converter, I use the STM32 microcontroller's ADC.

  • 0 in reply to Behtan Umut Mercan
    Guru 140200 points

    Hi Behtan,

    I think it has to do with the input bias current of OPA659 and the leakage current of D2 (1N4148). If the input bias current of OPA659 is higher than the leakage current of D2, then the input stage of OPA659 saturates and a wrong output voltage can be seen.

    This can be simulated by using the 1N4148 and BAV99: TINA-TI assumes a high saturation current for the 1N4148 (1nA) but a very low saturation current for the BAV99 (1.1fA):

    behtan_opa659.TSC

    This is a really nasty trap when designing a peak detector!

    A remedy is to choose rectifying diodes with leakage currents being much higher than the input bias current of buffer OPAmp, small signal Schottky diodes for instance. Or to mount a high ohmic resistor in parallel to the C1 / R3 combination:

    One last word: Please remove R1. Inside the feedback loop such a resistor is no isolation resistor but erodes the phase margin. It does more harm than good.

    Kai 

  • 0 in reply to kai klaas69
    Prodigy 90 points

    Hi Kai,

    Thank you for the suggestions. I'll try all of them. 

    Do you think that my first opamp(U1)  needs an isolation resistor/compensation circuit to drive 1nF capacitive load?

    (I think adding 22M res will increase its stability but I am not sure whether U1 needs an external component.)

  • 0 in reply to Behtan Umut Mercan
    Guru 140200 points

    Hi Behtan,

    R3 is its isolation resistor Relaxed

    Kai