LCD Display

Discussion in 'The Projects Forum' started by fredp, Jul 14, 2010.

  1. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    Hi, I started an earlier post about an OP-amp, but I wanted to start at the beginning.


    I want to scale an LCD display linearly with current in a wire so that it 1 A corresponds to a displayed value of 20, 2 corresponds to 40, etc. I have a 0.001Ω shunt, that the current goes through, so I can measure the voltage across the shunt.

    The first problem I ran into is that the voltage across the shunt is so small that you cannot really use a scaling circuit and get a meaningful voltage value. You would need to scale the 0.001 V to 0.0002 V and I don't think this makes sense. I was attempting to amplify the shunt voltage, but that results in amplification of the noise, and I cannot find anyone who can help me adjust the amplifiers offset.

    The only other route I can think of is to use a higher resistance shunt, but then the power dissipated at the shunt is very large. I don't think that is a good idea. I want to have a range up to 100 A.

    Is there any simple way to achieve this?
     
  2. tom66

    Senior Member

    May 9, 2009
    2,613
    214
    You might want to consider a hall effect sensor IC. These use magnetics to detect current flow, have a very low resistance ~0.001 ohms or less, are very accurate and much cheaper...

    I've seen 100A models, and I think you can go higher.

    Not to be confused by actual hall effect sensors which detect the presence of a magnetic field.
     
  3. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    Tom. How does that help scale the display? I can already sense the current.
     
  4. tom66

    Senior Member

    May 9, 2009
    2,613
    214
    Such a sensor produces an output in volts, instead of millivolts, so you can avoid scaling altogether.

    If you're at the noise floor, there's little you can do. You can try low noise op-amps. Maxim IC make some good ones.
     
    fredp likes this.
  5. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    I guess it is helpful if the output is in volts--although I'm sure some scaling would be required to get the 20 mV/amp display.

    Is there no such thing as an LCD meter that you can scale the display? It seems like this should be that difficult. The powersupply has a current meter on it. I wonder how that one works.
     
  6. tom66

    Senior Member

    May 9, 2009
    2,613
    214
    The power supply will use a 0.01 ohm shunt, because those are very common.

    At 3 amps, it drops 30mV, which is easy to amplify (although if they're using a typical op amp, which is very likely, the input offset is ~1mV, so your actual reading is 3.1 amps...)

    This output will then probably be sent to a microcontroller which performs the scaling using a multiply operation.

    If I assume a current flow of 3 amps... the drop is 3mV.

    3mV is way too low to amplify using commodity op amps. At this point, the offset voltage becomes very large compared to the input and you are just amplifying noise. As I said, you will need a low noise op amp. You will also need a VERY good power supply with good filtering, the PSRR (power supply rejection ratio) also plays an important part.

    But if you know you are going to be continually dealing with very high current, i.e. at the 100 amp mark, then the input offset plays a small part only and you could get away with using a normal op amp.
     
  7. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    Tom,

    I have been unable to find a 0.010 Ω shunt from any supplier. The shunt would need to be pretty large because it will dissipate 100 W of heat at 100 A. If I could find such a shunt, it would solve my problem. I have the scaling working already.
     
  8. tom66

    Senior Member

    May 9, 2009
    2,613
    214
    So you got it working?

    Of course a 0.01 ohm shunt is not a solution. Even your 0.001 ohm shunt dissipates 10 watts!

    That why I was suggesting a hall effect current sense IC.

    Although these may dissipate even 10 watts of heat as a loss across the chip's internals, at least they don't have too much problem remaining in specification with such an amount of heat because the hall effect sensor is not very temperature sensitive.
     
  9. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    Tom,

    I have been unable to find a 0.010 Ω shunt from any supplier. The shunt would need to be pretty large because it will dissipate 100 W of heat at 100 A. If I could find such a shunt, it would solve my problem. I have the scaling working already.
     
  10. fredp

    Thread Starter Active Member

    Sep 24, 2009
    41
    0
    Well, the OP amp method with the 0.001 Ω shunt does seem to work fairly well. It is able to scale the mV signal by 10. Unfortunately there is a 30mV offset on the output of the OP amp when I configure it as an non inverting amplifier with the appropriate resistors.

    I have tried all day to get help from this forum on removing the offset, but know one seems to be willing to offer info on this issue. References on op amps never mention this, so I have given up on this route.
     
  11. tom66

    Senior Member

    May 9, 2009
    2,613
    214
    I'm surprised no one has told you... it's really quite simple.

    There is an offset nulling pin on some op amps. Connect a potentiometer, wiper to offset, one pin to V- and the other pin to GND. Then adjust the pot until you get zero offset, or thereabouts.

    Disadvantage of this is it's not perfect (i.e. temperature differences will increase the resistance of the potentiometer, causing the offset to drift slightly, and unless you use a trimpot you could accidently change it giving you incorrect readings.) Also you still have the problem of noise to contend with. Amplifying small signals is difficult, because noise gets amplified along with it.
     
  12. bc108

    New Member

    Jul 18, 2010
    10
    2
Loading...