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.

OPA4188: Output distortion

Part Number: OPA4188
Other Parts Discussed in Thread: OPA188, TINA-TI, OPA4192, OPA2189, OPA189, OPA4170, OPA4191, OPA191, OPA170

Hi,

I've designed an input signal conditioning circuit using OPA4188 as shown in the figure below:

The input signal is a sine wave with amplitude that ranges from 0 to 14 Vpp and frequency that ranges from 1.5 kHz to 10 kHz.

I simulated the circuit using LTSpice and it performed as expected. However, when I implemented the circuit and tested it, I found that it presents an output distortion, as shown in the figure below:

The figure above shows the results when the input frequency is 10 kHz. However, this distortion is present for frequencies as low as 3 kHz. For lower frequencies the circuit does not present this distortion. I've repeated this test on each of the OPA4188 channels, on different boards and on different OPA4188s. They all presented the same issue.

I've seen similar questions on the forum (like OPA4188: Strange sine distortion and OPA188: TI precision lab EVM OPA188 Output distortion ) but none of them reaches a conclusion.

Did anyone reach a conclusion on these other questions? What could be causing this distortion and how can I avoid it?

Regards,

Filipe

  • Hi Filipe,

    increase a bit the amplitude and the frequency in your simulation and TINA-TI will clearly show you visible output distortion!

    You should not expect a perfect large signal performance from an operational amplifier with a unity gain bandwidth of 2MHz, a slew rate of 0.8V/µsec and a settling time of terrible 20µs...

    Kai
  • Hi Filipe,

    This distortion appears to be an interaction between the slew-boost circuitry in the OPA188 and the high input and feedback resistances used in your circuit. Testing in the lab showed that reducing the feedback and input resistances to around 15k decreased the severity of the distortion significantly. 

    Another option would be to substitute another op amp here. While it's not available in a quad version, the OPA189 does not exhibit the same distortion in this application. The dual version, OPA2189 is in preview right now, and we're expecting samples in late April. If a conventional, non-chopper op amp could work for your application, you might consider OPA4192 as a replacement. 

  • Alex,

    Thanks for your reply.

    I was suspecting it had something to do with the OPA188's output stage and the high resistances. However, due to the characteristics of the sensor this circuit interfaces with, this high input resistance is required. Therefore, I think I'll have to change the op amp.

    I am already using OPA4192 in other circuits in this project, but it has a much higher supply current consumption, which made me use OPA4188 in this circuit. I'm also using OPA4170. Do you know if the OPA4170 has the same output stage as OPA4188? If not, I'll try replacing OPA4188 for the OPA4170, given that it consumes 10 times less current than OPA4192.

    Regards,

    Filipe

  • Hi Kai,

    The input signal is limited to 10 kHz. I've tested the same circuit with another op amp, which has a unity gain bandwidth of 785 kHz, a slew rate of 0.4 V/usec and it worked perfectly, without any distortion.
    The distortion seems to have something to do with the OPA4188's output stage, as Alex replied below.
  • Hi Felipe,

    yes, Alex is right. The distortion decreases if the input and feedback resistors are decreased.

    Kai
  • Hi Filipe,

    OPA4170 should be fine. You might also want to take a look at OPA4191, which is a lower power version of the OPA4192.

    This may not be an output stage issue. Slew boost circuitry is typically added to the input stages of the op amp to improve step recovery in lower bandwidth op amps - when the inputs of the op amp begin to pull apart, the slew boost circuit turns on to increase the slew rate of the input stage, then turns off as the inputs come together again. Figure 51 from the OPA191 datasheet shows a typical block diagram of a slew-boosted op amp. 

     

    In this case, it's possible the larger feedback resistors, combined with the op amp's input capacitance causes enough of a delay for the slew boost to start acting up. This may be exaggerated due to lower stability (again from large resistors and parasitic capacitance). I'd actually recommend adding 22nF capacitors in parallel with R228 and R223 to improve the circuit's stability.

    I've attached two TINA-TI simulations of the proposed modification. One models the overall transfer function of the circuit, while the other is a stability check. I used the OPA191 macromodel for this simulation, but you can substitute the OPA170 macromodel if you want to test with that. 

    We also have a TI Precision Labs series on stability, the first segment of that can be found here. This series goes into more detail on how to investigate possible stability issues in simulation. 

    TINA-TI Files: 

    Transfer Function

    Stability Analysis

  • Alex,

    I've just tested the circuit replacing OPA4188 for the OPA4170. It performed as expected, without any distortion. I tested it with input frequencies as high as 40 kHz (4 times higher than the actual input maximum) and didn't get any distortion. I'll just have to check if the OPA4170's higher Vos will be an issue. If it is, maybe OPA4191 will be the best option.

    Thanks for the simulation files and the link for the TI Precision Labs series. I'll use them to verify the final circuit's stability.

    Best regards,

    Filipe

  • Glad this helped! I'll go ahead and mark this thread as closed, but if you have any other questions, feel free to either reply below (which should re-open this thread), or create a new question on E2E.