Near-ground op amp sensing

Discussion in 'General Electronics Chat' started by Shagas, Jan 30, 2014.

  1. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    Hello

    I'm currently trying to make a current sensor using a 0.01 ohm resistor and an op amp with gain across the resistor.
    Here is the schematic:

    Csense.png

    It says in the datasheet of the LM392 that it is suitable for sensing 'near or at ground' but I'm not getting any results.

    With about 130 ma flowing through the rsense I get a 1.7 mV (should be 1.3mV) drop across the rsense if I measure it with a meter but the op amp gives me some -3mV on the output.
    I also tried flipping the op amp inputs across the rsense around and I get a similar negative result.
    Am I overlooking something obivous and/or important here?

    Thanks in advance
     
  2. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    Ahh , I think I found the problem.
    The op amp ground was a few mV off in respect to the Rsense ground due to the current flowing in the supply rails.
     
  3. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    What voltage are you using vcc???
     
  4. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    5V powering the op amp and 5V powering the load.

    The problem is that I'm getting some strange results.
    With 1.8mV drop over the rsense I am getting a ~550mV output on the LM392N
    which is wierd because with a gain of 101 (as shown in the schematic) I should be getting an approximate output of 180mV , not 550mV .
     
  5. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    perhaps you should go to one of the older dual supply op amps? they work to below ground.
     
  6. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    Have you tried increase the rail voltage? ??
     
  7. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    I've seen some of those but I'm constrained to what I have in my parts box at the moment.

    I just tried changing to a TS272CN (output can swing to ground)
    and I'm getting a result of 250mV on the output which is.... well somewhat closer to the calculated result but still about 60mV off.
    Is it an offset issue or something like that ?
    I haven't gone into the very details of oP amps yet so maybe someone can help me out here.
     
  8. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    What effect would that have?
    Also I'm constrained with a 5V supply atm . I am building myself an ATX extension :)
     
  9. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    I could be wrong but I thought the rail voltage had to be higher than the input voltage ..
     
  10. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    Yes it does but if you look closely , the non inverting input is actually almost at ground.
    It only rises by a few milivots when there is current flowing through the 0.01 ohm resistor.
     
  11. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    The circuit you posted in #1 will work provided:

    a. The Vss pin is tied to the same point as the bottom end of the current sense resistor. (as you found out)
    b. The voltage drop across the current sense resistor has to be >> than the input offset voltage of the opamp. (likely not)
    c. The gain of the opamp is lowered. (100 is too high for b. above)
    d. the opamp is capable of differencing all the way to Vss or slightly lower. (Yours is ok)
    e. the output pin of the opamp is capable of pulling all the way to to Vss.(?)

    I have done similar using an LM358 with a 10K pull-down resistor between the output pin on the opamp and Vss in order to allow it to pull closer to 0V
     
  12. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    Thanks alot for your answer Mike


    So how do I solve the problem of the input offset voltage? Do I just look for an OpAmp with a smaller input offset?
    I don't mind if the whole result is offset by some value because I will be reading the output of the OpAmp by a micro anyway so I can take care of the offset in software.
    What is important though is for the output of the OpAmp to be linear across a reasonable range
     
  13. crutschow

    Expert

    Mar 14, 2008
    12,974
    3,220
    Check the offset value of the op amps. You are likely seeing the effects of that voltage. To sense 1mV you need an op amp with much less than a 1mV offset. Or you need to add an offset adjust network with a pot.
     
  14. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    To do it right, you need an instrumentation amp.

    Another possibility, use a high side current monitor like a ZXCT1009 and its bretheren.
     
  15. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    I'll look into an instrumentation amp but I prefer the idea of adjusting the offset.
    I'm guessing I need a negative voltage for that.
     
  16. Shagas

    Thread Starter Active Member

    May 13, 2013
    802
    74
    Ok so lets say I powered the Op-amp by +9 -9 Volts.
    Could someone advise how to adjust the offset with a pot?
    I looked online but I cannot find any circuits/tutorials for offset adjust for non-inverting configurations.
     
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The input offset voltage is multiplied by the gain of the circuit; in your case, 101. So, in the case where you were getting a 1.8mV drop across your load current sense resistor, the output from the opamp (with no input offset) should be around 181.8mV, but you say you're getting ~550mV. 550mV less 181.8mV leaves 368.2mV; which translates to a 3.645mV input offset when you divide by the circuit's gain of 101 - which is within the maximum input offset specifications for TI's LM272. See the datasheet on TI's website.

    ST Microelectronics has various specifications for the TS272; it depends on the suffix letters. As always, see the manufacturer's datasheet to obtain their specifications. In any case, the input offset voltage spec appears to be lower than that of the LM272.

    You can find out what your opamps' amplified input offset is by shorting the noninverting input (+) to ground, and measuring the output voltage.

    There are opamps that have zero input offsets, but they're generally rather pricey. There are also opamps that have means provided for adjusting the input offset.
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Oops, I guess I should've hit refresh before I replied. :rolleyes:
    Anyway, have a look at a better opamp, like an OP07. Digikey sells them for under a buck, even if you're buying them singly:
    http://www.digikey.com/product-detail/en/OP07CP/296-12853-5-ND/475983

    Input offset is typically 60uV (that's microvolts), which when amplified by a gain of 101, would be ~6.06mV. However, it has inputs to null the offset using a 20k pot; see the null circuit on page 5 of that datasheet.

    I'm sure there are other options, but my time (and bandwidth) is quite limited at the moment.
     
  19. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,231
  20. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,231
    Forgot the other question.
    Why such a low sense resistor if you working with ma why not 0.1 ohms?
     
Loading...