This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

REF3430: About Kelvin connection

Part Number: REF3430
Other Parts Discussed in Thread: REF3450, REF3440,

Dear TI


Is there any data on the REF34xx that shows how the IR voltage drop across the PCB trace changes with and without the Kelvin connection?

If data is available, will it be compensated?

Best Regards,


  • Hi Taroimo,

    The force/sense connect is a device feature that can be taken advantage of by customers when the VREF is far from the ADC.

    It also depends on the load current draw. The IR drop voltage = Load Current * Resistance from OUTF to Load.

    This is very system specific so is no practical graph we can provide.

  • Dear TI


    Thank you for your prompt reply.

    Excuse me, but I have one more question.

    As shown in the attached figure,

    I would like to connect a resistor of the order of kilo Ω used in the pressure sensor bridge circuit between Out_S and Out_F.

    Is this all right?

    Best Regards,


  • Hi Taroimo,

    The REF34xx will try to make the voltage at OUT_S = VREF. The REF34 might be able to support close to 0.5V upwards on OUT_F to adjust for the drop across the top "R" with a maximum of 10mA.

    What is the purpose of this circuit? 

  • Dear TI


    I am very grateful to have the specific values ​​revealed.

    This circuit belongs to the customer, so I can't clarify what it is for now.

    But if there is no problem, I would like to add it.

    By the way, why is there a limit of "OUT_F = OUT_S + 0.5V"?

    I thought that OUT_F and OUT_S would not be a problem if they did not violate the absolute maximum ratings.

    Best Regards,


  • In my opinion it looks like the customer wants to create a precision current sink with one device to bias a sensor. This cannot be supported if it violates ABS Max or RoC.

    OUT_S is a sense input and OUT_F is the output voltage of the REF34. The ABS max output voltage of the REF34 is 5.5V, so if the 5V variant is used, only a max of 0.5V can be used for adjustment. So the maximum voltage drop across "R" can only be 0.5V for the 5V variant.

    The force/sense is meant to be used for smaller IR drops across traces. For lower voltages I recommend to keep a similar requirement in mind. For lower voltage variants, 0.7V can be acceptable.

  • Hi Marcoo Z,

    Thanks for your advice.

    I understood that "only a max of 0.5V" refers to the case with REF3450.

    For example, when using REF3430 or REF3440, the allowable voltage drop is 0.7V, isn't it?

    Why is the former not 2.5V and the latter 1.5V?

    Best Regards,



  • The force and sense output is connected to the output buffer in a combination like the one below. The Op Amp will buffer the output to make sure the positive and negative voltages are the same. The OUT_S terminal will be regulated to equal the bandgap voltage. This is why OUT_S has to be the bandgap voltage.

    OUT_F will always be larger than OUT_S. The limit to how large OUT_F can be while still maintaining some regulation is limited and I gave it in the previous post. Keep in mind that if OUT_F is significantly larger than OUT_S the regulation will be impacted.

  • Hi Marcoo Z

    Thank you for your explanation.

    I understood that this is to maintain output regulation.

    Also, please tell us the following two points as supplements so far.

    The first is why VOUT(OUT_F) is independent of input voltage and is defined as a maximum rating of 5.5V.

    Is it because the drive voltage of the output stage buffer is stepped down to nearly 5.5V due to IC protection, etc.?

    The second is that the maximum ratings of the input voltage is stated as "Vref + 0.05V".

    It can be interpreted that it is necessary to fill “VIN≧0.05V” even when the IC is not operating.

    Isn't it safe to think that this refers only to the operation of the IC?

    I am sorry for the many questions, but We would appreciate if you could cooperate with us for a little longer.

    Best Regards,


  • Hi Taroimo,

    I am glad to help where I can.

    Let me look into these questions. I currently do not have the answer.

  • Hi Marcoo Z

    I'm grateful for your cooperation on the matter.
    We look forward to hearing from you.

    Best Regards,

  • Hi Taroimo,

    The ABS max is an error and it will be fixed in the next revision that is coming up.

    The 5.5V ABS max is a limit with the internal designs and I cannot go into more detail into it.

    Keep in mind that the specifications on the EC table for the REF34 assume the OUT_F and OUT_S are shorted together. Any large deviations on OUT_F can deviate the specifications from the EC table since the OUT_F deviation is mean to be in the mV as per section