Instrumentation Amplifier Keeps failing

Discussion in 'The Projects Forum' started by msrav1, Sep 28, 2008.

  1. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Hi,

    I am having problems using INA118 instrumentation amplifier, I keep killing it. I am feeding a 10mv, 10Hz signal from a Signal generator, the dual supply is coming from 2 9V batteries connected together. I am currently referencing the INA118 to the common voltage of the 2 9v batteries, but A kind person before mentioned that I am not referencing the INA118 properly because the signal generator may have any DC voltage, and that could be my problem.
    I am confused as to how i should reference it?

    (IN my schematic the INA118 is the one labelled opamp6)
    Any Help would be greatly appreciated, Thank You
     
    Last edited: Sep 28, 2008
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    This thread is virtually a duplicate of what is posted in this thread:
    http://forum.allaboutcircuits.com/showthread.php?t=14615

    Please do not create duplicate threads for the same topic, as it makes for confusion.

    But try connecting a 10MEG resistor to the V1 side of R1, and another 10MEG resitor to the V1 side of R2, and the other ends of those two resistors to the INA118's ground. That won't eliminate static electricity, but it will help a great deal to ensure that the INA118's input voltage limits aren't exceeded for too long.

    Better yet, tie the ground references together.
     
  3. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Sorry for the similar post, i was sort of really desperate for help, it wont happen again,

    Thanks for the advice. But by putting those resistors will it not affect the come mode rejection ability of the amplifier?

    I was also thinking that to overcome this problem, instead of using a signal generator that is connected to mains, maybe i could use a portable signal generator (acutally it is a portable ECG emulator, runs of a 9V battery). Wouldnt the DC value of the portable signal generator be at a lower, safer DC voltage? If so I guess i can avoid using those 10megaOhm resistors?

    The INA118 has 40V input protections, so am i killing the INA becuase the DC voltag of the generator relative to the INA's ref voltage is greater than 40V? Or does it have a different common mode rejection maximum? I cant seem to work it out from the Datsheet


    Thank You
     
  4. pebe

    AAC Fanatic!

    Oct 11, 2004
    628
    3
    Your inputs are floating. I assume that your signal generator gives an unbalanced output, with no series capacitor.

    So if you connect the amp ref to the inverting input (as well as the battery junction) you will stop it floating, and your output will be in the same phase as the input.
     
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Right now the inputs are basically floating to whatever arbitrarily high level they might reach.

    You could try connecting one side of the input or the other to the INA118's reference ground via a large resistor, or both sides via a resistive divider network. But somehow, you need to make certain that there is some common reference. Otherwise, you're going to continue to fry instrumentation amps.

    If it were an ordinary situation, I'd suggest clamping diodes - but those will always be drawing some current, even at levels below Vf, and below Vr.
     
  6. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    If you are using an oscope to measure the signals then you may have a sneak ground path through the scope probe ground clip.

    You might be able to use a battery and a potentiometer along with a resistor attenuator to create a floating voltage source. This way you will be able to use your oscope without introducing a sneak ground path.

    hgmjr
     
  7. pebe

    AAC Fanatic!

    Oct 11, 2004
    628
    3
    But the sneak ground path will almost certainly have an AC component which you couldn't remove with a DC potential divider. I would suggest a thick wire to bond the two instruments grounds together and then connecting their grounds to the amplifier ref point, the centre tap of the batteries, and the inverting input of the amplfier. No resistors are needed.
     
  8. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Hi,

    I tried connecting the inverting input to the Vref on my new INA, but after using it for 30mins and then powering off, and then powering back on about 30mins later, it doesnt work anymore, it keeps going to the the positive rail. Even when i switch the inputs, ie make the -ve input higher, it still goes to the +ve rail.
    When i get my new INA tomorrow I will try to do the resistor divider suggestion; connect 10megaohm resistors from the inputs to the Vref.

    Its really worrying me now, have got a lot to do, but cant even get the INA to work properly.
     
  9. pebe

    AAC Fanatic!

    Oct 11, 2004
    628
    3
    As long as ref is also connected to batt centre tap, what you have done shouldn't have caused a fault. While the op-amp is still in its faulty condition, I suggest you do some voltage measurements of -ip, +ip, op, and ref, relative to batt-.

    Then it might be possible to see what's gone wrong.
     
  10. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Hi,

    I have the -ve input and the common/centre voltage of the 2 9V batteries connected to vref.

    Input voltage: 10hz sine wave 0.344 Vrms
    -ve input: 0V relative to Vref
    +ve input: 0.315 Vrms relative to Vref
    output: 8.67 volts relative to Vref

    Relative to -ve Supply from battery:
    -ve input: AC reading: 0V, DC reading:+8.43 V
    +ve input: AC reading: 0.315 Vrms, DC reading: +8.25V
    Vref: 8.42 V (DC)
    output: 16.41V (DC) (0V ac reading)

    The gain is only set to 2.
    What could have happened? Could it be some other problem rather than the floating voltage problem?

    Help is very much appreciated, thank you kindly.
     
  11. pebe

    AAC Fanatic!

    Oct 11, 2004
    628
    3
    Baffling!
    I see you have a gain of 2. So you have a 50K resistor between pins1 & 8?
    Could there be a leak between pins1&2?

    What is the effect of connecting the two input pins together? Does o/p voltage then equal ref voltage?
     
  12. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Hi,
    It seems it was a breadboard problem aswell. On a different breadboard, I used a OPA2604 (not the INA118 inst-amp which i need to use) and followed Pebe's advice of connecting the -ve input and common of batteries together to form the common for the circuit (this opamp has no vref). The circuit worked fine, the output was 1.5Vrms (input .3Vrms, gain set to 5).
    When i transferred the circuit to the exact same pins of my original breadboard, the output was 0vrms and the DC reading was 6.5 V. Clearly there is something wrong with the breadboard. When I moved the OPA2604 back to the fault free breadboard, it worked fine, the old bread board did not kill it.

    However there is a weired pattern to this problem, here is a sequence of events that occured:
    Last week (unawares of the floating voltage problem), my first INA118 failed in Breadboard(no.1). I got my 2nd INA118 and used it in Breadboard(no.1), it did not work, so i used a different breadboard(no.2), it worked there. However after operating the INA118 for about 20mins, powering off, then coming back to it, the INA failed. This is when i posted my problem on the website. After receiving advice from the kind people of this forum, I used a better referencing methodology. With this methodology I operated my new INA for 30mins, powered off, came back to it 30mins later, and the INA did not work anymore, I used it on another breadboard and still did not work. Then I did what i mentioned above (used a opa2604 with Pebe's referencing methodology i was following for the INA in a new breadboard (no.3), it worked fine, I moved it into breadboard (no.2) it didnt work, moved it back in breadboard 3 it worked fine, i did this several times, the results were the same, and the OPA hasnt died out).

    I just dont know what to do next, is my circuit still wrong? or is it a bad brand of breadboard I am using? I am totally bedazzled.

    Thank You.
     
  13. pebe

    AAC Fanatic!

    Oct 11, 2004
    628
    3
    I didn’t realise you were breadboarding the circuit. That is most likely to be your problem. The input impedance of the INA118 is 10¹º ohms and you are feeding the +ve input through 33MΩ, so the slightest electrical leak across the board on to the +ve input will upset readings.

    For test purposes, use a high quality IC holder and solder the input components and the gain resistor directly to its pins. Then complete the rest of the wiring, fit the amp into the holder and turn it upside down so that the IC holder pins are not in contact with anything else.

    Then see what the results are.
     
  14. msrav1

    Thread Starter Member

    May 26, 2008
    10
    0
    Hi,

    I thank everyone for helping me get to the issue of the problem, especially Pebe, thanks a lot. I used a better quality breadboard earlier today and it works fine, but I will take no further chances and do the soldering as you said, hopefully it will be all ok now.

    I hope one day when I become a competant Electical engineer, I can help others like how you guys have helped me, these forums are great, thank you.
     
Loading...