Post Go back to editing

ADM4073H Current sensor, RS+ & RS- input resistance? can anyone help?

is there any info on the input resistance of the RS+ & RS-? (RG1 & RG2)

or any notes of how to filter any noise into RS+ & RS-?

I am trying to add and "RC" filter to reduce the noise into the ADM4073H

When I add the RC = 1k & 0.1uf, the ADM4073H gives the wrong output current

    1k resistor from Rsense to RS+ and 1k resistor to RS-

    0.1uf cap between RS+ & RS-

I am measuring the output current with a amp meter and the readings from the ADM4073 do not match,
they are as follows:  amp meter = 4A, ADM4073 = 500mA (with the 1k resistor)

if I bypass the 1k resistors the readings are very similar.

What I need is an example of a filter circuit that can be added to the Rsense circuit (between RS= & RS-) to reduce the input noise

any comments will be appreciated

Thank you

Fausto bartra

  • Hi Fausto,

    You cannot add any resistance between Rsense and RS+ or RS- because that will affect the gain of the high-side current sense circuit, which will lead to the severely reduced output voltage that you've seen. Here's the math:

      The ADM4073 family's voltage gain is set by internal resistors: Av = RGD/ RG1

      For ADM4073H, the ratio is set at 100 to achieve 100V/V gain. Since RGD is given as 12kOhm, RG1 must be 120 Ohm.

      When you add 1kOhm between Rsense and RS+, it looks like it is added in series to RG1, and the resulting gain is now

             Av = RGD / (RG1 + 1k) = 12k / 1.12k = 10.7V/V

      This is far lower than the 100V/V expected, hence the output reading 500mA while the real current is more like 4-5A.

      When you bypass the 1k's, you short them out again and RG1 returns to the original value, so the gain returns to 100V/V as well.

    What are you trying to accomplish with input filters at RS+ and RS-? Since this is a high side current sense, the input is usually a supply rail, and that should be well-bypassed with a wide range of capacitor sizes to shunt noise away. If your supply rail is very noisy and not bypassed well, there are bigger problems for the rest of the circuit that is powered by it.

    Please let me know what your original goal was, and I can try to help you fix it in a way that doesn't affect gain too.

    Best regards,

    Catherine

  • Catherine:

    Thank you for your reply

    Because my readins are all 1/10th of what they should be, I was trying to add some filtering

    See below the measurments:

    Amp meter DC        ADM4073H Vout DC

             1.8A                                     144mV

             5.5A                                     503mV

             9.2A                                     1042mV

    The above readings are average from six different boards

    I am using the ADM4073H (100V/V)

    Rsense = 0.003 ohms

    Also notice that I have connected the RS+ and Vcc pins of the ADM4073H to get it to work

    Any suggestions will be appreciated

    Thank you

    Fausto Bartra