LTC4359 Reverse Polarity Protection for low power load

Hi there,

I'm attempting to use the LTC4359 to provide Reverse Polarity and Reverse Current Protection for the input of a DCDC converter that charges a battery. At the moment, I have spiced and prototyped something similar to Figure 10 in the datasheet. Please see the spice image below:

The DCDC converter, which is "off," is modeled as two diodes pointing towards the LTC4359 (body diodes of the MOSFETs). When the battery is present, the Reverse Polarity and Reverse Current Protection both work fine. But I run into a problem when the battery is not connected AND the input to the DCDC converter has been connected in reverse. In this case, Vout levels off around -0.7V (I believe ~two diode drops below ground). I have confirmed this in spice and on the bench. My concern is the bulk input capacitors on the DCDC converter are aluminum electrolytic, which are polarized and could be damaged by the negative voltage. Is there any way to prevent this.

I also confirmed it is indeed the body diodes that clamp the voltage to -0.7V, removing one lets the voltage wander even lower. 

In the end, I'll need to combine this solution with that of Figure 11 from the datasheet to handle input voltages between 15V and 110V. If there is a better option please let me know.

Any advice is greatly appreciated.


  • 0
    •  Analog Employees 
    on Oct 5, 2021 7:53 AM

    Hi Phillip,

    Would you kindly post here your whole schematic including the DC-DC converter so we could help you better. We're having difficulty in trying to visualize how the two diodes (or mosfets) represent the DC-DC converter and what kind of DC-DC converter it is. We also have a lot of converters that you can simulate in LTSpice along with your reverse polarity protection circuit since using diodes may not properly behave the same as a converter.

  • Thanks for getting back to me. The DCDC converter is built using discrete components. The second "diode" is the body diode of a constant current controller downstream (M3 below). Here is a picture that shows the LTspice test fixture with the following changes:

    1. DCDC converter and Constant Current Controller added.

    2. 1K resistor replaced with diode (in preparation for higher input voltages).

    3. Source modified to drop from 0V to -20V.

    (The battery V2 is in the schematic, but currently not connected). Technically the body diodes of M2 and M3 are in parallel in this case.

    Simulation shows Vout levels off around -650mV, which I have also seen with prototypes. If M1 is removed, Vout wanders even lower.

  • 0
    •  Analog Employees 
    on Oct 12, 2021 12:47 AM in reply to prkarls1234

    Hi Phillip,

    We suggest trying to simulate each stage of your design separately before combining them. Following a design from the datasheet for the LTC4359 is a good first step. Then, make sure your buck converter is also working as intended before combining it with your input protection circuit to ensure proper simulation. We noticed that your buck converter seems incomplete, the M1 and M2 gate pins are just connected to their respective source pins. When your buck converter is working, then you may try adding a proper constant current controller if it's really needed.