• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 47 subscribers
  • Views 445 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

INA116 output peaking

Donald Le Roy
Prodigy 120 points
Other Parts Discussed in Thread: INA116

I am trying to use the INA116  for high impedance measurements in a frequency range of 0.001 Hz to 10 kHz with the amplifier configured for G=1 (ultimately need to be able to handle 10V peak up to 10 kHz using +/-15V power supplies).  The output of the INA116 is driving a 1 meter 50 ohm cable into the input of an EG&G 7260 Lock-In Amplifier which has a 10M ohm || 30 pF  input.   I am observing peaking at the output of the INA116 when being driven by a low impedance (50 ohm) sine wave at 1 V RMS (1.414V peak) starting at about 3 kHz but do not see the peaking at the guard outputs indicating an issue with the composite instrumentation amplifier.  According to the datasheet this should be well below the full power bandwidth for the device (based on the slew rate that would be 18 kHz at 10V peak however this is only specified for G=10 to G=200 not for G=1).  Below are some measurements at both 10 kHz and 100 kHz (I realize that at a G=1 there is peak of ~ 5 dB at ~ 500 kHz but is shown as flat to 100 kHz) - all inputs voltage levels are RMS:

@ 10 kHz
    100 mV input 0dB peak
    250 mV input 0dB peak
    500 mV input 0.2dB peak
    750 mV input 0.4 dB peak
    1000 mV input 0.42 dB peak
@ 100 kHz
    100 mV input 0.3dB peak
    250 mV input 4.2dB peak
    500 mV input 8.1 dB peak
    750 mV input 6.1 dB peak
    1000 mV input 3.7 dB peak - probably exceeds the SR of the INA116

Above measurements limited to 1V RMS due to LIA maximum input sensitivity.

I am less concerned with the 100 kHz measurements but the peak at 10 kHz is troubling (there is also about a 4 degree phase shift which exceeds the expected phase shift due to the 800 kHz  G=1 bandwidth which should be about 0.7 degrees - ARCTAN [10/800]).

I have tried using the TINA 3 op-amp model provided by Matt Hann in an earlier post (since there is no SPICE model for the INA116) and do not see the peaking or phase shift issues (modeled the load as the 10 M ohm || 30 pF for the LIA input and a 100 pF shunt capacitance for the 1 meter cable. Is it possible that the amplifier does not like driving the cable capacitance and if so would a series output resistor be needed or helpful (datasheet lists amplifier Load Capacitance Stability as 1000 pF)?  If that is not the issue Is there a way to compensate the amplifier to get flat amplitude and the expected phase due to bandwidth limiting at 10 kHz?

Any insight would be appreciated.


Don

  • 0
    TI__Genius 12335 points

    Don,

    It is quite possible that the cable capacitance is causing a degradation in the phase margin of the output amplifier which could cause peaking and so a series resistor is worth a shot. However, instrumentation amplifiers are also highly sensitive to parasitic capacitance at the gain resistor pins (pin 1 and 16 on the INA116). If your ground plane extends around these pins it may introduce enough parasitic capacitance to causing peaking in the transfer function. 

  • 0 in reply to John Caldwell
    Prodigy 120 points

    John -

    Thanks for the response.  I have been quite careful to keep ground and power supplies away from the amplifier inputs and the 2 Rg input pins so I don;t see that as an issue.  I certainly intend to try adding a series resistor somewhere between 250 and 500 ohms to the output.  I did notice that there was at least 1 other post similar to mine which did not appear to have a solution (output slew rate was suggested, but that doesn't seem to be the case here).

    I don't have the boards available to me at the moment as they are at the customer's site and committed to testing (up to 1 kHz) to meet a grant deadline of the end of July.  Once I get the boards back I will try the resistor addition.

    Since the datasheet does not specify slew rate for G=1, is that number available or is it  not specified?

    Any other suggestions would be appreciated.

    Don

  • 0 in reply to John Caldwell
    Prodigy 120 points

    So I have tried adding a series resistor to the amplifier's output (250 ohms and 1.0 k ohm) with no effect on the output peaking.

    I am attaching a scope shot of the input and output of the INA116 with 1 VRMS from a 50 ohm source applied at 20 kHz with the amplifier having +/-15V supplies.  According to the datasheet this should be within the output amplifier specifications.  Based on this image the output of the amplifier appears distorted (slew rate limited?).  NOTE - since it was suggested that parasitic capacitance at the gain resistor pins can cause peaking pins 1 and 16 were lifted from the board as a precaution (with no effect).

    Again I will pose the question on original response to your post - is there a specification for the slew rate at unity gain?  The datasheet only lists slew rate at gains of 10 to 200.

    Any additional insights would be appreciated.