• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 687 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.

INA2126 output is not expected

Gijs De vries
Prodigy 30 points
Other Parts Discussed in Thread: INA2126, INA126

Dear all,

This is my first time on this support. I am a dutch student.

I have a INA2126 instrumentation amplifier used to amplify the results of a load cell. I supplied the amplifier with 5VDC and I have a reference voltage of 2.5VDC. The load cell uses a wheatstone bridge with two outputs that I feed into the INA2126. Typical output voltage of the load cell is 2.31VDC (both outputs). I have a gain of approximately 450.  I have to research what happens when one of the two inputs get disconnected.

When I don't apply force to the load cell, the outcome of the amplifier is around 2.5VDC; which is to be expected. However, when I disconnect one of the two sensor outputs using a toggle switch, my output is amplified to the wrong rail (so I think).

An example:

O+ = 2.31V (sensor output one)
O- = 2.31V (sensor output two)

O+ is connected to Vin-
O- is connected to Vin+

When I disconnect O+, making O+ 0V, the amplifier output is pulled down to the negative rail (0V). However, I do notice that the output goes to the positive rail first and after a few millisecondes the output is pulled down to the negative rail.

When I disconnect O- instead of O+ than the effect is opposite. The output is first pulled down to the negative rail and after a few milliseconds the output is pulled up to the positive rail.

This is to me the exact opposite of what should happen. The output equation is (Vin+ - Vin-)*Gain + 2.5 ( (O- - O+)*Gain + 2.5 ). When O+ is disconnected, Vin- goes to 0V and the output should be amplified to the positive rail. When O- is disconnected, Vin+ goes to 0V and the output should be amplified to the negative rail. As my example above demonstrates, the real outcome is the opposite of what I would expect.

I found an article of a Texas instruments employee who defines the input common mode voltage range. He explains that the amplifier will not operate linearly when outside the input common mode voltage range. The input common mode voltage is given by (Vin+ + Vin-)/2. When both sensor outputs are connected, the input common mode voltage is 2.31VDC. When one of the two sensor outputs is disconnected, the input common mode voltage should be 2.31/2=1.155. In the INA2126 datasheet, i see the graph for the input common mode voltage range and its corresponding output. An input common mode voltage should be greater than 1VDC. When disconnecting one sensor output, I do approach the minimum of the range but I should stay within the correct range. 

Article: http://www.planetanalog.com/document.asp?doc_id=528175#msgs

I am having a real trouble understanding why this occurs. Do I exceed the input common mode voltage range which would explain non-linear behaviour? Or is there something else that I am missing?

Hope to hear from you guys soon.

Thank you in advance.

Kind regards,

Gijs de Vries
Student

  • 0
    Prodigy 30 points

    **EDIT 21-03-2016**
    When I reconnect the sensor output, I see the same behavior.

    Reconnecting O+, Vin- returns to 2.31V, the output of the amplifier first goes to the positive rail (5V) and then returns to its equilibrium of 2.5V. When reconnecting O+ the time between the output being near the positive rail and equilibrium is approximately 1/10th of the time the output is near the positive rail when disconnecting O+ (Output goes to negative rail).

    Reconnecting O-, Vin+ returns to 2.31V, the output of the amplifier first goes to the negative rail (0V) and then returns to its equilibrium of 2.5V. When reconnecting O- the time between the output being near the negative rail and equilibrium is approximately 1/10th of the time the output is near the negative rail when disconnecting O- (Output goes to positive rail).

    **EDIT 21-03-2016**

    The first response of the amplifier (positive rail when disconnecting O+ and negative rail when disconnecting O-) is the response I would expect. This is also approximately the discharge time of my capacitor between sensor output and ground. When reconnecting the sensor output, the time it takes from the opposite rail to the equilibrium is approximately 1/4th or 1/5th of the discharge time.

  • 0
    TI__Mastermind 25795 points

    Hello Gijs,

    Here is a list of useful topics...you may find #9 helpful.

    e2e.ti.com/.../design_notes

    The transfer function of the INA126 assumes that a known voltage is applied to each input. The device's operation is not characterized with floating inputs. When one of the inputs of the INA126 is disconnected, the common-mode voltage is unknown. This is because the common-mode voltage is the average voltage at the inputs of the device, and only 1 potential is connected. Put another way, the output of an instrumentation amplifier requires 3 pieces of information and only 2 are being supplied.

  • 0 in reply to Pete%20Semig
    Prodigy 30 points
    Dear Pete,

    Thank you very much for your quick response!

    Kind regards,

    Gijs de Vries