Current Shunt Test. Big errors are posible!

Discussion in 'General Electronics Chat' started by PeterCoxSmith, Mar 2, 2015.

  1. PeterCoxSmith

    Thread Starter Member

    Feb 23, 2015
    148
    38
  2. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    measureing current with a shunt does not require a 4 wire connection, measuring the resistance of shunts does, sinced they are usually low resistance, which should use a 4 wire to test. measuring the voltage drop across the shunt with a volt meter of high enough impedance would not require a 4 wire connection.
     
  3. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    Well, yes and no.

    Yes because you typically already have a 4-wire connection where the current goes thru large wires or conductors and a two-lead voltmeter is just applied to the sense resistor.

    No because you cannot just wire in the sense leads just anywhere but just at the sense resistor, unless you can live with some bad readings.
     
  4. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    the leads from the shunt to the meter movement are always connected to the shunt directly. and the meter is ocasionally an ampmeter, like a 5 ma shunt uses a 5 ma meter which at full scale is marked for the current rating of the shunt.
     
  5. PeterCoxSmith

    Thread Starter Member

    Feb 23, 2015
    148
    38
    yes that is fair comment. in any application you have to judge whether the connection resistance is significant or not.
     
  6. Reloadron

    Active Member

    Jan 15, 2015
    963
    232
    Overall I found it to be a good and informative writeup. Possibly yes, it could have been better illustrated but overall it got the point across in an easy to understand manner. All of the current shunts I have encountered have been 4 terminal devices with the true shunt resistance measured across the sense terminals.

    50 Amp Shunt.png

    They all look about the same. They all have 4 terminals.

    Shunts1.png

    Anyway, overall I saw it as a good and informative writeup.

    Just My Take
    Ron
     
  7. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    One need not use such a shunt to measure current. Most any resistor will do if it meets the resistance tollerance for the given current. Such is a frequent practice, especially inside a power supply PCB.

    Yes, care must be taken to separate the sense voltage from the current path, basically connecting as close as possible to the resistor itself.

    Aside: a 4 terminal measurement is also referred to as a "Kelvin" sensing. It's the same Lord Kelvin for whom the temperature unit is named for.
     
  8. Reloadron

    Active Member

    Jan 15, 2015
    963
    232
    Yes, I realize that and on a good number of circuit boards and in a good number of applications resistors are indeed used as current sensors. Just as long as we can live with the measurement uncertainty. Also, the resistor must meet the necessary power dissipation requirements. So I agree, you run a current through a resistance and measure the subsequent voltage drop. That said the focus of the article was the type of current shunts that I posted. I am also familiar with Kelvin measurements or more commonly called 4 wire measurements.

    The thing with current shunts is that we select a shunt for our intended application. Every aspect of the shunt is important including the alloy the shunt is made from. Manganin, which is a copper and manganese mix is commonly used for better shunts (I have no clue as to the $1.98 shunts off the boats from China) is generally the preferred alloy because of its temperature properties and good shunts also use bolts made from silicon bronze and again because of the temperature properties and to avoid thermal emf issues on the measurement plane. However, I saw no need to include this pile of information in the short article about shunts. Overall, as I mentioned, I found the article to be well written and informative but did feel a few more illustrations would have been in order.

    Ron
     
    PeterCoxSmith likes this.
  9. PeterCoxSmith

    Thread Starter Member

    Feb 23, 2015
    148
    38
    Thanks. If you buy a shunt generally they have 4 connections as you say. In some schemes a plain old resistor gets used with 2 connections and you have to make sure you pick off the sense in a , shall we call it, sensible place. Years ago I worked on a flight control system development. The manufacturing dept decided to rationalise the wiring, which was a good idea before production. The operations director arrived at my desk, weeks later, to tell me the performance had dropped right off. My first question was did you move the sense wires from the resistor. And of course they had, The sense wires were a pain to fit and route through the airframe.
     
  10. PeterCoxSmith

    Thread Starter Member

    Feb 23, 2015
    148
    38
    You make a lot of good points. I have been thinking about writing a book about sensing. This was just a short blog, and I'm experimenting with it. I'm writing it because a friend asked me to do it. It is hard to decide on the "value". Any design requires a lot of input not just one factor.
     
Loading...