Post Go back to editing

Ringing and Overshoot at 1 MHz frequency AD8648

Category: Hardware
Product Number: AD8648

Hello,

I am using AD8648 in a resistive load circuit where the input is a sinusoidal 2 VPP wave at 1 MHz, the dual supply is +3V and -3V. I firstly build the standard non-inverting amplifier circuit but it gave me high ringing and overshoot at the output, just to make sure the functioning of opamp, I switch the frequency to 1 KHz but still the ringing and overshoot was observed. I have also attached a 5pF compensating capacitor in series with the feedback resistor  but still the same problem. Can anybody please enlighten the problems that I am facing, any help would be deeply appreciated.

Thank you in advance

  • Hi,  ,

    Can I get a copy of your circuit so that I could check it? I tried to simulate it using your given configuration and here is what I am getting, I cannot replicate the ringing and overshoot. 

    Thank you!

    Regards,

    Mae

  • Hello, thank you very much...I will attach the circuit but I think the problem is that in simulation it works fine but in practical it does not work.

  • Hi,

    I think you should reconsider the supply voltages because this amplifier’s voltage range is between 2.7V and 5.5V. Your ±3V voltage range reaches the amplifier’s absolute maximum value.
    „The device is NOT designed to work at the values contained in the ABSOLUTE MAXIMUM RATINGS table.” – you can read more about it on this page. 

    I think the main problem is caused by the way the signal is measured. Since you are measuring the output with an oscilloscope, you are putting a capacitive load on it, which can cause oscillation. You should try putting a 100 Ohm resistor in front of the probe.

    Connecting a capacitor to the inverting input may cause instability in the circuit. This problem can be solved by connecting a capacitor in parallel with the feedback resistor.

    I hope any of these solutions will help you!

  • Thank you very much for the message, I decreased the input voltage to  ±1V but still I measured an identical signal disturbances. The 100 ohm resistor that you are talking about should be in series with the probe right? so basically the resistor and the probe capacitance will be series? please answer so that I can try the circuit. 

    the capacitor Cx mentioned in the drawing is the input capictance of the opamp but actually in the actual circuit the capacitor is connected in parallel to the feedback resistor.

  •  Dear  ,

    as an extra I would like to add a few thoughts on observations regarding your problem.

    a)  is completely right. Connecting the capacitor to the inverting output just worsens the chances of oscillation. And of course, you should never power your IC close to the absolute maximum ratings. Using 6 volts as power could have already destroyed your amplifier.

    b) These spikes at the output should not be called ringing neither overshoot. Ringing usually is something that has periodicity, overshoot is tipically related to control processes with feedback and setpoint.

    c) You should never connect the power through crocodile clips. One wrong move, a slight touch can destroy your circuit.

    d) It seems like you have connected an LM324's output to the input of your circuit. LM324 has a crossover (zero-crossing) distortion which probably caused the spikes on your signals.

    e) You should also check the input signals of your circuit. The crossover distortion is probably already there at the input of your circuit.

    f) I'm not sure but it seems like there is an inductance at the output of your amplifier, which can explain how can you get +- 5 volts spikes while using only +-3 volts Vcc.

    g) For simulating Analog Devices ICs, you should use LTSpice. You can find the AD8648 model in LTSpice which models many details related to oscillation and frequency dependence.

    Regards, Dénes

  • Hello, thank you for your message. I have decreased the dual input supply to 1V but still observing the spikes. LM324 circuit output is completely isolated so that should be fine. I will replace the crocodile clips with some steady connectors. Capacitor is not connected in the inverting  input, the actual circuit is as follows:

    could you please tell me about the capacitance loading that Nemagsa talked about, I think this could be the problem....

  • Simply add a resistor directly to the output of the amplifier as Nemagsa suggested.

  • thank you...I will try that. one more question...is it important to make the ground of the supply and the input signal same or it can be different. I read in the datasheet that it is good to have different grounds for the supply and the input signal

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.