Post Go back to editing

Added feedback circuit for LT3623 as negative regulator

Dear ADI/LT experts:

I need to control the output of the LT3623 (configured as a negative regulator) with a positive voltage.

When this IC is configured for a positive output, this is easy because I only have to put the desired voltage into the ISET pin. When it's configured as a negative regulator, it's not that straight forward.

Can you please help me figuring out if the following circuit is correct? Do you see any possible risk in this implementation?

U3 is inverting the desired voltage (4 v) and U2 is doing the difference between the output of the regulator and this voltage. R2 (6 k) is to add a 0.3 V margin to the desired voltage.

The counterpart of this is that I need to provide a separate -5 V rail. The power sequence will be the following:

  • Apply VIN to LTC3623 with the RUN pin at 0 V (disabled).
  • Power the OPAMPs U2 & U3, with V3 at 0 V
  • Enable the RUN pin of LTC3623
  • Turn on V3 (variable supply).

Another concern is that I'm using the output voltage as the reference for Imon. Can it be a problem? The circuit was tested in LTspice and it works as expected. Please note that the spice model of the IC was updated recently.

Thanks for your help!


  • Dear team:

    I built the circuit by using the DC2131A development kit and it's working. The only difference is that I've added an additional 24.9k resistor between the output of U2 and the output of the regulator (PGND).

    It's working as intended but I can see that the pin IMON is outside the specification:

    As it's referenced to the output, I have always negative voltages on IMON pin. It's connected with a 10k resistor to the output of the regulator (PGND pin). Can this be a problem or it is OK?

    Thanks in advance

  • Hi Ody,

    This absolute maximum rating should be from IMON to PGND of IC. Do not violate the negative voltage rating at any time between these pins. A -0.3V gives me idea there may be ESD protection diode between pins that you don't want to forward bias by going above -0.3V else you may render IC incapable. 

  • Thanks Fil for your answer! Then in my application this pin is working within specs.