Instrumentation amplifier

Discussion in 'The Projects Forum' started by galvanicdude, Sep 23, 2011.

  1. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Hello,
    I am trying to build a usable ECG for some music project.
    By scouting round Internet, most projects seem to be using an instrumentation amplifier.

    So far I have no form of analogue filtering, as just want to make sure the instrumentation amplifier is working correctly. However, after applying considerable digital filtering I still can't monitor the peeks analogous to heart contraction.
    From a previous experiment (connecting the electrodes to a simple gain stage then connecting them to my sound card, recording, and filtering) I was able to monitor the wanted signal just by applying digital filtering.

    So I used this schematics
    I have R=100k , Rgain = 470k potentiometer.
    And am using 741s opamps.

    The main problem for the moment seems to be that I'm getting a negative voltage at the output of my instrumentation amplifier. What I'd like is voltage varying between 0v- 5v.

    If anyone could help the inexperienced me, that would be great.
    thank you.
     
  2. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    Can you describe the voltage output you get. Also the 741 opamp is an old relic. Of all the opamps you have selected this one of the worst you could select:eek::eek:
    What are your supply voltage for the 741 opamps. One thing that is also important. Are the ECG electrodes used. Are you using proper electrodes
     
  3. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    The output voltage is currently -3.80V. It doesn't vary when I change the value of the Rgain potentiometer, nor does it change when I disconnect the electrodes (something I have just realised)...

    I'm using 741s as I know people who have used them on similar projects in the past and worked. I am not trying to get an extremely precise ECG, so I guess they will do?

    I'm using electrodes that marketed them selves as being proper electrodes. But I am skeptical to how important they are as people have used cut out beer cans to get a usable ECG reading.
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    What are your supply voltages?

    You should be using ±8v to ±15v.

    Please post the circuit as you have it configured now.
     
  5. tgotwalt1158

    Member

    Feb 28, 2011
    111
    18
  6. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    My 741 are powered with ±9V.

    The schematics is exactly this one.
    I have all resistors equal to 100K
    The Rgain is a 470k potentiometer.

    Also, I have read that an IA (Instrumentation amplifier) has the characteristic of inverting the output. This would explain why I have a negative current on the output.
    What is a easy way round that, knowing that I need the output to vary within 0V and 5V?
     
  7. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Thanks, I came across that in another thread (or forum). This would be a solution, but then I would need to get new components, which is bit silly as I'm sure with the ones I've got here i can make an ECG.

    I've basically done what this guy has done. However, my version doesn't seem to work.
     
  8. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    If you want posetive electrode output swap the electrode cables. You should also know that ECG electrodes will dry out very quicly after the bag is opened. The must be kept in an small air tight container. And yes they are important in order to get a stable signal. It would also help if you could give us a picture of how your signals looks. You use a PC and sound card so that should not be to hard.
     
  9. tgotwalt1158

    Member

    Feb 28, 2011
    111
    18
    Ok! Remove the variable pot for gain since its value is too high, calculate the value of R-gain, and put a fixed resistor instead, if it works then well and good, later you could add a 500 ohm pot for fine tuning the gain.
     
  10. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Pretty sure that is not the problem as I get a negative value wherever I position the electrodes.
    Swapping them, turning them upside down doesn't change the output voltage, my problem is elsewhere.

    They're well house kept. I've got electrodes that are reusable if rinsed! So for the moment I'm not worried about them. I'm looking forward to the point at which I'll be looking for a stable signal, for the moment I have no signal at all.

    I am not using a sound card anymore. When I was, I had no problems.
    I am now using an Arduino Uno board. Which is why I need a voltage varying between 0V and 5V. I don't, thus, I'm not able to visualise at the moment (well I'm lying, I've just got a strait line at 0V), when I can I'll post the waveform.

    Kewl, I'll try that, when done I'll let you know where I'm at.

    Thanks for your helps!
     
  11. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Ok so I recalculated Rgain by using the formula:
    G= 1+ (2 x Rref)/Rgain.
    I got it from this site

    Right so supposedly the voltage resulting from the contraction of our heart is in the realms of 10mV. Thus, I need a gain of 100 or more?
    For G = 100, knowing Rref = 100K
    I need a Rgain to be minuscule?
    Or am I being really stupid (probably am..)

    Anyways, I put Rref = 500Ω. Still same result...
    I've checked my circuit several times, I can see any unwanted contacts.
    Any other ideas?
     
  12. tgotwalt1158

    Member

    Feb 28, 2011
    111
    18
    You need to calculate R-gain and gain from the same design of the circuit which you are using, otherwise you may get wrong values for your components. Also, double check all of your wiring and connections which might be causing out put to go to the negative value. It may be due to a very trivial over looked error. Good luck!
     
  13. tgotwalt1158

    Member

    Feb 28, 2011
    111
    18
    I tried your formula. You told your IA should provide an out put of 0~5 Volt and that you want to amplify a supposedly 10 mv signal up to a range of 0~5 Volt out put. Considering maximum required gain to get amplification of the signal from 10 mv to 5 volt, gain would be 500, which by putting in your gain formula :G= 1+ (2 x Rref)/Rgain, gives R-gain=401 ohms, keeping R-ref at 100K. So my idea was correct that 450K pot is way high which you are using for adjusting R-gain.
     
  14. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Yup, It is way to high.
    But I have now tried with a 500Ω≈ resistor, there was no change.
    I also tried several different resistors and haven't gotten any satisfying results. I think my problem is elsewhere.
    I'm just going to take it apart and rebuild it. Hopefully, it will work this time.
     
  15. tgotwalt1158

    Member

    Feb 28, 2011
    111
    18
    There is also a chance that your 741 may have got damaged during of all these experimentation. So, a better idea woud be to try with a fresh piece of 741. Good luck!
     
  16. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It's really tough to build your own instrumentation amplifier. You need to use 0.1% tolerance metal film resistors, and even then you can't match the CMRR of a factory-made amp. They're laser trimmed on the wafer level for the best possible performance.

    Have a read through at least the first few pages of this Application Note:
    http://www.intersil.com/data/an/an1298.pdf
     
  17. galvanicdude

    Thread Starter New Member

    Sep 23, 2011
    8
    0
    Hello again,
    Thanks I read a bit of that document.
    And have ordered a factory built instrumentation amplifier recommended for what I need. INA121?

    In the meantime, I'm having problems with opamps in general..
    I don't seem to understand them. I've read about the theoretical side of them but when I apply what I think is theoretically right, it doesn't work...

    My main problem seems to be having output signals pegged at 3/4 of the voltage supply?
    (This happens on 741s and LM358s)
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It will help us to help you if you can post schematics of the circuits that you are having problems with.

    Please keep in mind that the 741 is ancient, and needs at least a 9v supply for you to see any kind of output at all; and even then it will be in a narrow range. It doesn't do very well with a single supply. It really needs a ±15v supply; but better yet, use a more modern opamp.

    The LM358 can be powered by a single supply, but you'll have confusing results if you let the input signal drop below the ground rail.
    Datasheet: http://www.national.com/ds/LM/LM158.pdf
     
  19. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    Reverse the polarity of the inputs on the third op-amp.

    This answer may be far too late to be of any use except for educating you about how to invert the output when you need to.
     
Loading...