LF347 as instrumentation Amplifier

Discussion in 'General Electronics Chat' started by viju, Oct 16, 2008.

  1. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    Can LF347 OP-Amp be used for gain more than 100?I intend to use this IC as an instrumentation amplifier to amplify sensor signal.
     
  2. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    Do you have the data sheet? It should have all the information you need in it. Burr-Brown (now part of TI) makes very good IA's. They would have superior performance because the internal resistors are trimmed more accurately.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    You really don't want to try to build an instrumentation amplifier using an opamp.

    Please read Intersil's Application Note 1298; it's a great intro to instrumentation amplifiers.
    Direct link: http://www.intersil.com/data/an/an1298.pdf
    Check out INA114, INA118, and INA128, formerly Burr-Brown, now TI.
    Very good and easy to use instrumentation amplifiers for a great price.
     
  4. DickCappels

    Moderator

    Aug 21, 2008
    2,660
    632
    Yes, you can, but at 1 kHz, you will only have 40 db of loop gain. It depends on what you are trying to do.

    Back in the old days, we made instrument amps from the op amps of the days. They worked for our purposes, and some of those applications called for precision performance. AND WE LIKED IT.

    And we didn't take the bus to school...we walked through 4 foot drifts of snow in our bare feet. AND WE LIKED IT.

    And we didn't buy hamburgers at Mc Donalds...we would corner a cow in the pasture and... AND WE LIKED IT.

    So, today, as yesterday, you can buy dedicated instrumentation amps, or you can roll your own. It just depends on what kind of cost/performance you are shooting for.
     
  5. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You lost me on the loop gain... Can you explain how you got that? Mind you, I'm not taking sides on make vs buy.
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK, I will ;)
    With an INA128 or similar, you can wind up having great CMRR, low offset, low drift, and just the gain you need with just three external components; a resistor or pot to set the gain, and bypass caps for +V/-V to ground. Minimal PCB real estate, minimal parts count.

    There's not much hope to even approach that CMRR, offset, drift and stability using an LF347 and one or two dozen external components, even if the resistors were .01% tolerance.

    Sure, an INA114, INA118, INA128 etc. costs several times what an LF347 does. But in this case, you really do get what you pay for - even for just one or two prototypes.
     
  7. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    As you suggested I tried to procure INA 114.But not available in my town.Instead shopkeeper gave me an equivalent IC - AD 620.I have powered this IC with +/- 12 Volts and a pot of 500 ohms for gain.But I am confused with the reference terminal. I have a common point of +/- 12 volts and 0 of signal.Will I have to connect the reference terminal to these common. Rightnow I am not getting the linearity.You have adviced to connect the capacitors in the power source terminals. Will 0.1 micro do?

    How to set Zero offset and span?

    Wookie:

    I am also running the op-amp circuit that you suggested sometime back. I want to go further deep into that.Let me learn from your suggestion.
     
  8. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    That reference level is the reference for the signal. 0 volts is usually good for the reference.
     
  9. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Viju, are you providing an input bias current return path? Your input signal(s) must be connected to ground through some path.
     
  10. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    Yet to resolve the linearity issue. I have connected the pin 5 to the ground which is my PSU's ground. I also tried connecting to "0" of input signal.Please help/
     
  11. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You need to post a complete schematic, including how your signal source is connected, and your power supplies.
     
  12. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    According to the datasheet, a 499 Ohm resistor is needed for a gain of 100.
    It should be a precision metal film resistor.

    You also need to tell us what your load resistance is.
     
  13. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    Pls find the attached schematic. Till now I am measuring with DVM.Also the gain resistor is a variable trimmer pot set at 499 ohms.I have tied PS ground,reference terminal and input signal "0".

    Wookie:
    Once it isthrough I will connect this to my PLC which has >10K impedence.
     
  14. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It looks like you have grounded the inverting input. Not good!
    Put a 10k resistor between the inverting input and ground. That should be enough to provide input bias current.
     
  15. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You have me baffled. How does grounding one input cause a bias current problem? How will a series resistor remedy that "problem"?
    Grounding one input doesn't prevent the amplifier from amplifying. It does defeat input common mode rejection, which is the main reason for using an inamp.
    Viju, what kind of sensor are you using?
     
  16. mindmapper

    Active Member

    Aug 17, 2008
    34
    0
    The datasheet page 14 - 18 deal with the input circuit and reference connection. You will find lot of important information there. Usally the zero is adjusted on the tranceducer and span is the gain adjustment of the instrumentationamplifier. But that is depending of the kind of sensor you're actually using.

    Grounding the input circuit could be the reason of unlinearity if the sensor has to be floating refered to gnd.
     
    Last edited: Oct 21, 2008
  17. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    I tried with the method suggest by Wookie.

    Its works fine with gain upto 5. But I need gain for 100 and 1000.I have attached my complete schematic here.Please suggest me where I am going wrong.

    Wookie:

    My constant current source works well as you suggested. Here I have used it for the resistance measurement( RL1).Resistance value will be in mΩ.

    Please help me to resolve the issue.
     
  18. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Have you read what the datasheet says about the importance of proper grounding? Pay particular attention to Figs. 46 and 48.
    In your schematic, R4 is shorted out, and does nothing, as I pointed out in my previous post.
     
  19. viju

    Thread Starter Senior Member

    Sep 28, 2008
    120
    0
    Yes I have gone through the datasheet.I have used a SMPS for powering the IC.I have connected all ground points( as shown in the schematic) to the powersupply common.Also I tried with R4 removed.

    System works well for signal greater than 1 volt.(upto gain 5) I get excellent linearity.

    Once when I connect with the signal level as shown in the schematic I get the following results.

    Input signal
    1mV
    2mV
    3mV
    4mV
    5mV
    Output Voltage
    2V
    3V
    5V
    6V
    7V


    These results are for gain 1000.

    I too have connected the reference pin to the power supply common as mentioned in fig,46 & 48. I have SMPS with +12,-12 and a common point. To this common point I have connected the reference terminal( pin 5).
     
    Last edited: Oct 22, 2008
  20. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Can you post a photograph of your test setup?
     
Loading...