Differential Op Amp Question

Discussion in 'General Electronics Chat' started by rainyday101, Feb 14, 2012.

  1. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    I have a hall effect sensor that is going to read 0 to 25 amps. The hall effect sensor has an output of Vs/2. I plan on using 5 Vdc for VS. This gives a null offset output of 2.5 Vdc at 0 amps of current. The sensor sensitivity is 37mA/1A. At 25 amps this would give a sensor output of 3.425 Vdc. Per my attached schematic, I am trying to offset the 2.5 Vdc with U1 to get a 0 Vdc output, and then adjust gain to get a 0 to 5 Vdc output. U2 is being used as a buffer and the 0 to 5 Vdc wil then be used as an analog input on my Arduino. I have learned/refreshed myself alot on op amps in the last 5 hours and cannot get this to work/an not totally sure what I am doing. I can never get the output at 0 or even close when I have a 2.5 Vdc input on U1 (0 amps from the sensor). I keep ending up with 2.5 Vdc output at U2 and U1 in this case. Any help would be appreciated.
     
  2. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    Will get back to you tonite unless someone beats me to it--the best circuit is simpler and easier to understand--you are not far off, but the way you are dealing with the offset will not work. It appears that you are using something like the Allegro ACS710KLATR-25CB-T--is that correct?
     
    Last edited: Feb 14, 2012
  3. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    The amp was a MCP660 by Micro Chip. Low power CMOS, single supply.
     
  4. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    Do you have -5V available?
     
  5. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    This is the simplest way to do it. By making the offset pot 1K, its Zint ≈ less than 250Ω so that its adjustment will not affect gain appreciably. On the other hand, a small resistance change to the 440K resistor will affect offset, so it must be dialed in carefully. Interaction can be avoided by changing the 440K to 100K and adding a 2nd non-inverting amp with Av of 2. 7.

    Tricky task to do without a negative power supply.
     
  6. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    Which op amp should I use for this? LM324? 741?
     
  7. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    I can have -5Vdc available if needed. The circuit is being made from scratch.
     
  8. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    Early on, I thought that it would need a -5V supply, but it worked out OK without it.

    The 741 or 324 op amps will not do because their output does not go lower than 50mV, nor higher than about Vcc-1.5V.
    It needs a rail to rail output op amp--the MCP660 should do OK, but 32V/us is overkill.
    If selecting a new rail to rail op amp, go for about 0.5V/us or so.
     
  9. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    I just figured out the 741 or 324 would not work in Tina TI. I will look through their spice models and see what they have that is rail to rail and try your circuit again. Another thing that I tried was just a diff. amp with 2.5Vdc going in on V1 and the hall effect sensor voltage going in on V2. The hall effect output is 2.5 to 3.425 Vdc @ 37mV per amp. This gave me the diff Vdc of 0 to .925 Vdc on the output (V2-V1), which would just need to be scaled by another amp to get 0-5 vdc.
     
  10. Brownout

    Well-Known Member

    Jan 10, 2012
    2,375
    998
    No. If the output is 37mA/1A, then at 25 amps, the output is 925mA. It looks to me this is throwing the whole scheme off, thinking the output is a voltage that's proportional to current. But the spec you've quoted says it's actually a proportional current. This will have an effect on the architecture.
     
  11. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    The output is .925 Vdc + the 2.5 VDC null offset of the sensor. With no current the output of the sensor is at 2.5 Vdc, with 25 amps the output is 3.425 Vdc. (2.5 + .925)
     
  12. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    I assumed that this is a misprint since a constant current transducer (such as a Bell transducer) does not have a V/2 quiescent output. It seems very close to the 38mV/A Allegro transducer that I referenced.
     
  13. rainyday101

    Thread Starter Member

    Sep 24, 2009
    50
    0
    JimKeith,
    Just ran the simulation of your circuit and it works great! Thanks for your input.
     
Loading...