Deriving power from 4-20mA current loop

Discussion in 'The Projects Forum' started by GirishC, Feb 25, 2016.

  1. GirishC

    Thread Starter Active Member

    Jan 23, 2009
    58
    0
    Hi,

    I have designed a circuit which will be powered by 4-20mA current loop. As per my guesstimate I do not need more power at 3.3V and hence this should suffice. The circuit is based on shunt regulator TLV431B as shown below: TLV431B.JPG

    I have not shown a Micro connection but the Output of Opamp U3C (Pin # 7) goes to the ADC input of the Micro.

    1. Based on the the datasheet the min Vref required at TLV431B is 1.234V. With this potential divider values I set that.
    The output goes to 4.2V for any value of current between 4 to 20mA both inclusive. The I disconnected U2, U3 from the connection so only TLV431 is on the board but output remains unchanged.
    2. The datasheet also refers to value of Vk at Ik = 10mA. With this at 4mA the output voltage is 2.5V and at 20mA it is 4.5V.
    3. I replaced TLV and resistor divider by LM385-2.5 and the circuit work just fine. But then the micro needs 3.3V.

    Please suggest what I could be potential problem?

    Regards,

    Girish
     
  2. Dodgydave

    Distinguished Member

    Jun 22, 2012
    4,962
    743
    Are you trying to develop a 3v supply using a 4-20mA supply loop, or are you just wanting to measure the current using a micro?
     
  3. Papabravo

    Expert

    Feb 24, 2006
    10,135
    1,786
    Suppose the power supply to the current loop was 50VDC. Would you still get 3.3 Volts at the indicated point? I've never seen anybody put a regulator in a current loop, that's not really the purpose.
     
  4. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,761
    1,099
    U3 needs a minimum 5V supply, according to the datasheet.
     
  5. GirishC

    Thread Starter Active Member

    Jan 23, 2009
    58
    0
    @Dodgydave : Yes, I am trying to generate 3.3V and measure current as well.
     
  6. GirishC

    Thread Starter Active Member

    Jan 23, 2009
    58
    0
     
  7. GirishC

    Thread Starter Active Member

    Jan 23, 2009
    58
    0
    @Papabravo : Well to answer your questions there are two thing to understand
    1. The input SMJ will protect it beyond 33V
    2. The source is current controlled so the voltage is function of this current.

    Hence it will not hurt the shunt regulator.

    @Alec_t : Input is always higher than 5V.

    Interestingly I replaced TLV431 by LM431 and I get output 3.3V. But LM consumes more current than TLV according to datasheet.
    It appears that TLV needs more current but datasheet fails to substantiate it and I am hence confused.
     
  8. atferrari

    AAC Fanatic!

    Jan 6, 2004
    2,644
    759
    The way I understand you, whatever you do to generate additional signals ( keep in mind you said wanting to get power from the loop) will increase the current. At that moment, whatever that loop indicates, will be not the correct value.

    Are you trying to make a conversion current-voltage?
     
  9. Picbuster

    Member

    Dec 2, 2013
    373
    50
    To complicated. one chip is doing this for you look at xtr11X series chip but I <4mA
     
  10. Papabravo

    Expert

    Feb 24, 2006
    10,135
    1,786
    As I understand the way it is supposed to work you can't modify the current on the loop in any way if you are supposed to sense the current in the loop. You device won't play nice with the rest of the devices.
     
  11. ebeowulf17

    Active Member

    Aug 12, 2014
    678
    79
    Drawing current to power something within a current loop shouldn't be a problem. The transmitter regulates current in the whole loop, so as long as additional loads don't draw enough power to prevent the transmitter from driving current high enough, there's no problem (read about loop drop in the following paper.)

    https://www.google.com/url?sa=t&sou...1sc7hhAAItDNPeuSQ&sig2=_Z8OB4ioKpPmSVA6_k8OYQ

    Many current loop receivers are loop powered, so there's nothing theoretically wrong with attempting this. I'm not great with zeners and shunt regulators, so I can't comment much on the specifics of this circuit.

    One oddity though: if the goal is to measure current of the current loop and output it to the mcu, you need your reference resistor to be measuring the whole loop. R1 appears to be positioned to do this, but you're measuring voltage across R4. So it appears as though you're measuring the current that your added circuit draws, but not measuring what the current loop is doing. You've created a circuit that simply watches itself?!
     
Loading...