• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 762 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.

very low oscillation problem on LMP7731

cedric LEBOEUF
Prodigy 70 points
Other Parts Discussed in Thread: LMP7731, OPA376

Hello,

Using the LMP7731 with high gain (200), input resistor of 1k and feeback resistor 220k (with parallel capacitor of 4,7nF) to measure low amplitude 50Hz signal, we have very low oscillation problem on output around 10 millihertz. Replaced with OPA376, we didn't found these oscillations.

The main difference between the two is there is a current cancellation offset stage on the LMP7731 that doesn't exist on OPA376.

If there is an Ti engineer to help us to find the root cause of this very low oscillation that prevent us to use this amplifier on our board ?

Thank in advance

Best Regards.

 

  • 0
    TI__Guru 53996 points

    Hello Cedric,

    Hmm...That is strange. Can you show us the plots? Is it a "clean" , consistent cycle, or "noisy"? What is the amplitude? What is on the output?

    The OPA376 is a CMOS input, and the LMP7731 is a bipolar input. The OPA376 does not need bias current cancelling.

    Is the fixture thermally isolated? 100 seconds is too fast for air conditioning cycling, but is there anything else on the board that might be thermally cycling with a period around 2 minutes? The LMP7731 has input bias currents that, working against the 1K resistor, can vary with temperature adding to the apparent offset.

    If you want better noise performance, drop the resistors to 100ohms and 22k (+47nF), to reduce the effect of the current noise. The smaller you can make the 1k resistor, the better the noise performance.

    Regards,

  • 0 in reply to Paul Grohe
    Prodigy 70 points

    Hello Paul,

    I can't show the plots as the application is confidential. I just want to correct the frequency that is 0,2millihertz.

    We don't think that it comes from thermal cycling, because there is no problem with the OPA376. We are wondering if the LMP7731 is sensitive to external magnetic field of frequency 50Hz?

    Thanks for your reply.

    Regards. 

  • 0 in reply to cedric LEBOEUF
    TI__Guru 53996 points

    Hello Cedric,

    200uHz = 5000 seconds = 83 minutes.

    That could be room air conditioning cycling in a large room.

    We are not aware of any "low frequency" oscillations - or how they could even occur. Especially in a simple gain block application.

    What is the p-p value of the oscillation? Is the amplitude consistent or variable? Are the zero crossing point timing exact? Does it change over night?

    What happens if you wrap the device in bubble-wrap and seal it in a box? (poor mans thermal isolation chamber).

    The LMP7731 should be no more sensitive to external magnetic fields than any other device. Magnetic fields are usually picked up by external traces.

    Is there *anything* that would have a time constant of 5000 seconds (83 min) in the system? Heaters? Battery Charging? Relay? Are there two frequencies (say, switchers or clocks) that are running 200uHz apart - causing a VLF beat frequency?

    As I said, the LMP7731 has DC bias currents that could slightly change with temperature - causing a small baseline shift. The LMP7731 is also lower 1/f noise at low frequencies (<1Hz)  than the OPA376, so it could reveal small shifts that might be masked by the 376 noise.

    Normally when we see "low frequency" oscillations or cyclical shifts, it can be traced back to either thermally induced, or drifting supplies or references.

    Would you be willing to send the plots and/or schematic to me privately? If so, please send to paul.grohe@ti.com if you can.

    Regards,