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.

INA326: Filter configuration on the INA326 and replacing for the INA128

Part Number: INA326
Other Parts Discussed in Thread: INA128

Hi

Im currently using the INA326 for an EMG filter. I see that the suggested schematic has a filter in it. When trying to calculate the frequency it didn't correspond tho the 1kHz it says. Here I have a first question:

1) How the filter was calculated and how can I re calculate it to hava a cut-off frequency at 400Hz??

Question 2:

I was having a problem and the suggested answer was the lack of the path for the bias current. I put some resistors in it and the output remained at ~-800mV but this time it was stable. Im guessing this is normal no? to have an offset caused for the electrodes. Am I right? I used 47K, 1M and 10M and the behavior was the same...

I was thinking on changing the INA326 for the INA128. So here come questions 3 and 4

3)Does the filter at 1kHz can be applied to the INA128? I mean... can I use the same schematic as the one for the INA326?

4)I tried to do a High Pass filter to remove the offset by adding a capacitor in series with Rg (something like image below)

 But when measuring with the oscilloscope, the output signal still had the offset... Why it didn't work as a high pass filter? or did I miss something?

Thanks for the help! 

  • Hello Miguel,

    I will analyze this circuit for you with the INA128. I should have an answer for you in a couple of days.

    Best,

    Errol Leon
    Texas Instruments
    Precision Op Amp Applications
  • Thanks!!

    I already did some test and apparently the fact of using a resistor and a capacitor on the REF pin on the INA326 was because that component uses current instead of voltage. So the amplification is current based and that way, the current was transformed in voltage. Also the output filter start giving me some troubles because of impedance coupling.

    That I think would be interesting to put in an application note or something. 

    Im interested now on the behavior of the INA 128 with the capacitor in series with Rg

    Thanks a Lot!

  • Hi Miguel,

    Putting a capacitor in series with Rg does effectively create a high pass filter, but keep in mind the gain of the INA128 can never be less than one because of the 3 op amp topology it uses.  Therefore you can't use this method to block DC voltages. If you are looking for another way to AC couple your signal, you can put an integrator on your output and feed its output back to the reference node as shown in the attached simulation.ina128-ac-coupled2.TSC

    Regards,

    Zak Kaye

  • I was thinking on using a gain=22 and the capacitor in series to have a first order High pass filter. I test it couple of days ago and it worked fine...
    I could not open the file. Let me see if I can check it.
  • Yep... I cannot open that file... Is a compressed file? Can you send me a png orin a .rar file?
    Thanks!
  • It is a TINA TI simulation file. Here is a copy of the schematic:

    Best,

    Zak Kaye

  • But... that would act as a Low Pass filter and that does not solve the DC offset problem. I will attach my LTSpice simulation. I added the configuration of the INA128 with three opamps because of the lack of the INA component on the libraries (I have no idea of how to add new components)

    The green line is the output of the filters before going to the amplifier

    The blue one is the instrumentation amplifier output

    and the red one is after the Low Pass

    This one is the output from the instrumentation amplifier. 

    analog_filters.zip

  • Miguel,

    Yes an integrator is a low pass filter, but when you integrate the output and feed it back to the reference pin you are feeding back a phase inverted version of the output as negative feedback and this effectively creates a high pass filter. It will remove the DC component. I'm not really sure what to make of your simulations. I do not believe you have properly referred the output of your discrete instrumentation amplifier to the differential input to observe the true transfer function. You cannot achieve a gain less than 1 (0dB) with this topology.

    Regards,

    Zak Kaye

  • But how is a 0dB gain if the resistor is placed correctly? Im after all, amplifying the signal no? the differential input signal

    Regards
  • Hello Miguel,

     

    Thank you for your patience. The answer to your questions are below.

    1. How the filter was calculated and how can I recalculate it to have a cut-off frequency at 400Hz??
    • The filter is a passive low-pass with the following response fpole= 1/(2piRoCo). INA326's datasheet shows a way to achieve a 40dB/decade low pass response with the combination of the passive filter at the output Ro and Co with Pin5 R2 that is paralleled with C2. C2 influences the gain of the INA326 where at DC the capacitor acts as an "open" leaving the gain equation as G=2(R2/R1). As the frequency increases R2 decreases and eventually becomes a "short"; the result is a low-pass effect.
    1. I was having a problem and the suggested answer was the lack of the path for the bias current. I put some resistors in it and the output remained at ~-800mV but this time it was stable. Im guessing this is normal no? to have an offset caused for the electrodes. Am I right? I used 47K, 1M and 10M and the behavior was the same…
    • That is correct the resistors are needed create a path for the bias current.
    1. Does the filter at 1kHz can be applied to the INA128? I mean... can I use the same schematic as the one for the INA326?
    • The INA326 and INA128 internal structure are different. Where INA326 Gain is G=2(R2/R1) and INA128 G=1+(50kohm/Rg). INA128 does not have IAcommon (Pin5 on INA326). You can use post filter at the output with a Sallen-Key filter topology like you have in your schematic example. A second order low-pass will give you the desired 40dB/dec slope response.
    1. I tried to do a High Pass filter to remove the offset by adding a capacitor in series with Rg (something like image below)

     

    But when measuring with the oscilloscope, the output signal still had the offset... Why it didn't work as a high pass filter? or did I miss something?

    • With a three op amp INA topology adding a capacitor across Pin 1 and Pin8 at DC is the same as leaving it open (not populated). This will make the gain G=1+(50kohm/Rg) a gain of unity (1V/V).       The only way to ac couple and INA that I know of is having an integrator in the feedback (Zak proposed) where the output feeds into the integrator and the output of the integrator is connected to the REF pin.

     

    Below is the schematic of and AC response of INA128 with an integrator and Sallen-Key filter.

     

     

    If you want to have more attenuation at DC and lower frequency you can make a dual-slope integrator with a T-network in the integrator's feedback. This can be shown below.

    Keep in mind that this circuit has a gain of unity in the passband.  This can be scaled to desired gain by scaling gain in the sallen-key filter and the INA128.  Let me know if you have any questions.

    Best regards,

    Errol Leon

    Texas Instruments

    Precision Op Amp Applications