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.

selecting opamp for sensing piezo generated voltages?

Other Parts Discussed in Thread: THS3091, OPA192, OPA365, OPA301, OPA2301

Dear sir,

I am working on Structural Health monitoring project in which i need to sense the voltage generated by piezo.

This piezo has capacitance near about 5nF and it generates voltage waveforms of frequency around 500kHz in the range -5mV to 5mV. I need a opamp which should be able to sense this voltage and amplify it in the voltage range 0-3V.

I have tried this with THS3091 opamp but it won't be able to sense the voltages. Kindly help me to find a correct opamp with larger input impedance.

Thank you

  • Hi Mandar,
    I think you'd better off with a low noise, low offset, high input impedance front-end followed by a 2nd stage to get the desired output swing while preserving the bandwidth you need. I think the bandwidth you need along with the other requirements for high input impedance and low noise may be prohibitive in a single amplification stage.

    For the front end, I'd use a device like the OPA192 (5.5nV/RtHz, 20pA input bias current, 10MHz, RRIO, 25uV offset) to interface with your sensor, running at a gain of not more than 20V/V to preserve the 500kHz bandwidth. The 2nd stage would then need to run at a gain of 30V/V (or 15V/V depending on whether your input is 5mV or 10mV) to get the output swing you require. For the 2nd stage, I'd use a device like the OPA365 (RRIO, 50MHz, 25V/us, 65mA output current) to complete the design. The entire circuit can then run off a single 5V supply.

    I think you're also hinting that you need to shift your output so that -5mV input corresponds to 0V output. Right? For that, you may introduce a stable / well-controlled offset in the 2nd stage.

    Please review and update if I'm missing something.

  • Hi Mandar,

    Another possibility for lower noise may be to use OPA301 (3nV/RtHz, 150MHz, RR out, input CM voltage range to ground, 5pA max input bias current, 5mV max offset!) assuming that you have a means to correct / adjust for the large offset? With the higher speed of the OPA301, you may be able to get away with a single stage to handle up to 500kHz (but only barely as Open Loop gain plot shows less than 300kHz when operating with 50dB closed loop gain). You can definitely increase the gain on the front-end with this higher speed device, to maximize SNR and to minimize the 2nd stage noise impact.


    You may need to do some level shifting anyway to produce near 0V with -5mV from your sensor. This may be adopted to include correction for the offset voltage (assuming that's a possibility in your setup?). With the overall gain you need (~ 300x assuming 10mV signal translation to 3V output swing), if untreated this offset would overwhelm your signal. However, the OPA301 offset voltage drift (2.5uV/deg. C) is a lot higher than that of OPA192 (0.5uV/deg. C). So, you may have to consider the temperature range of operation when deciding.



  • Thank You, Hooman,

    Yes, this solution will probably work.

    And yes I was also thinking to amplify it in more than one stages.

    And You got everything what i want to do.

    But here I found a new problem. maybe my understanding of electronics circuits is not that great. probably you will be able to provide me a solution.

    In my system, i have a piezos at different locations bonded to an aircraft grade aluminum plate and we are sending a strain wave (a particular kind of tone burst signal) through one piezo. (Its like producing a wave in a material using expansion and contraction property of piezo with varying voltages ). this wave (tone burst signal)  travels through material and we want to sense it at other location. Depending on the nature of signal received we implement different algorithms to detect damages in the material.

    Now I am trying to develope a small scale instrument and an embeded system to do this job at very prelimanary level. so i need to sense this signal at sensory piezo. yesterday i was just tried one another opamp OP37 with gain of 2 noninverting configuration and checked a signals in an oscilloscope. But i dont know what is the problem with circuit it gets offsetted at a very high level like +5V and this offset changes continously containing the original signal. (even i didnt attached any offset voltage adjustment). I am attaching a schematic of the circuit. Even when i tap a signal at opamp input it was also offsetted. But when i remove the opamp and directly attach the piezo with oscilloscope the signal stays at 0V mean line (with -5mV to+5mV).

    Now one way i thought of this problem is like this, piezo which we are using also senses the evormental noise which is might be of low frequency and due to that low frequency waves are created which are producing ripples on which our signal get superimposed.  i have seen such a ripples in signal while analysing a signal directly with oscilloscope But that doesnt stays for longer and signal stablizes at 0V mean line. But why this large offset happens with opamp attached? maybe u will able to understand this.

    Again this op37 has Vos trim pins which i am not using (rather i dont understand what it actully do?)

    I am planing to use the instumentation amplifier circuit (with some manipulations in gain and offset) shown in application section of its datasheet as this opamp is able to sense the voltages from sensory piezo. (But I will do this when i get rid of this unusual offset and uderstand this Vos trim)

    maybe u can help me out with all this & why it is happening.

    Again, Is there any way so that maybe we can filter our signal of sensery piezo and then amplify it? because ripples due to envoronmental noise can lead to very high voltages in my case as gain will be around 300. (But they will get clipped at source voltage level).

    Again Thanks for Your valuable knowledge and support


    Mandar Kothavade

  • Hi Mandar,
    If you can, please post the waveform that you observe.

    I don't think you can operate OP37 at at gain of +2V/V as you've configured it. Datasheet shows Phase Margin of 71 deg. with Av= 5V/V.
    So, I'd recommend you lower the gain setting resistor (from inverting input to ground) to 2.5k or lower and try again.

    There are other low noise devices that are unity gain stable (no requirement for minimum operating gain). If that's what you need, I can send you a list.

  • Thank You Hooman,

    This is how my signal looks. Here yellow channel is the signal which I am sending using Actuator piezo and the blue channel is the signal which i am receiving from sensor piezo (blue signal Which is of our interest). This signal is around 175KHz. This signal in pic. shown below is without attaching opamp circuit. 

    After attaching opamp circuit the signal is same but it gets offset around +5V as i said earlier(i don't have a pic right now of this case). but in this case (without opamp circuit ) u can see that signals mean line is at 0V (look at the pointer of channel 2 blue at left indicating position of 0V ). 

    As your suggestion, I will try replacing 10K resistor with 2.5k or lower (but this will make gain of 5V/V). its ok for me I am trying these circuit just to find out that opamp I am using will be able to sense the signal from sensory piezo or not. 

    My final requirements are as I mentioned in the first post. I need to amplify the signal  (-5mV -- +5mV )  to a level of (0V -3V). and then this signal will be recorded by an ADC of my embeded system. Then we will further process it to get information about damages in material.

    Thanks for your support

    -Mandar Kothavade

  • Mandar,

    What supply voltages do you currently have available?
    Does one side of the sensor have to be grounded?
  • Dear Ron,

    As I mensioned before, I am testing this system. So currently i can provide any voltages in the range of +24V to-24V using regulated power supply.
    As soon as i will complete my testing phase. I have a plane to design a whole circuit and maybe i can include the regulated supply voltages into it as required. But that will have limitations because this system will run on two batteries of 12V each so i can provide +10V to -10V regulated power supplies in that case. And as per my knowledge i think this range of voltages will be sufficient to run what i am planning.

    About sensors, yes One side of sensor is grounded because of its placement.(I mean as we are pasting piezo on metallic plate. so its -ve electrode is kind of in contact with plate and so that plate needs to be grounded in order to sense and excite the piezos.)

    Thank You.
    And let me know if you need any other information.

    -Mandar Kothavade
  • Mandar,

    Building from Hooman's OPA301 suggestion. I recommend a dual OPA301 which is the OPA2301.

    To keep output in the 0V to 3V range, I used a 3V single supply. It is OK to change this to 3.3V or 5V if you wish.

    Each amplifier has the same gain 17.32 V/V or 24.77 dB. It will be easy to add more filtering if desired.