Schematic question - capacitor suggestions

Poley said:

Would I be able to check that the recommended use case for this is to be the very first circuity that the power supply should go to when entering the board?

Yes, you can use the LTC4368 as an e-fuse to your system.

Poley said:

If so, would you suggest adding more capacitance to either side of the component?

I would add capacitance to the output of the LTC4368. This way, the LTC4368 can control the inrush current going into the cap. This will prevent the input from sagging and creating an undervoltage fault, and also limit the SOA stress on the NFETs.

Poley said:

I am currenty having issues where my MCU resets when the PCB closes a 4.3A contatcor so I was wondering if there was anything that could be done to this circuit to bulk up to robustness.

I assume the MCU is downstream of the LTC4368. 
You should get a scope capture on the LTC4368's input, output, and gate waveforms.
This will give you a better idea of what the issue is.

Hey, thanks for the response!

So add more capacitance to the section with 100uF and 4.7uF? Or somewhere else?

Yes the MCU is downstream, I will try and get a scope into the product to see if I can see anything drop out. This is the basic stream that I am using:

From your diagram, it looks like you suddenly step the load (Rincon Contactor), and the LTC4368 hits its 10A circuit breaker threshold.
You can verify by probing the cooldown waveform on the LTC4368's RETRY pin.
If the system is restarting every 1.3sec, that's the issue.
You can add more output cap to the LTC4368 until the problem goes away (just make to scale the GATE cap with the output cap).

Here's more detailed troubleshooting tips:
Increase or disable current limit on the LTC4368 by reducing the sense resistor.
Use a current probe to see how far over the 10A limit you go when stepping the load, and for how long.
Start sizing up Cout and Cgate and recheck the waveforms until problem resolved.
If you are only briefly exceeding current limit, we can add an RC across the sense resistor to introduce delay into the overcurrent fault.

Hey, sorry I should have been more clear. The 10A is the just the device max current setting. The actual max current of the contactor is 4.3A for a max of 300ms. What is strange is that this only cuts out once then works when it reattempts on startup. It’s like something has to warm up before it is ok.

We need some waveforms to check if the LTC4368 is to blame for a brownout or overcurrent fault.

Hi Ashapiro , do you suggest any capacitance on VIN of the IC which might protect against quick dropouts of the supply? So maybe adding 2 150uF caps as shown here by the 2 C's? (Just 50V version instead of 25V)

Don't add Cin.
A big part of the LTC4368 is its hotswap functionality: inrush control into Cout when Vin is stepped. If you add capacitance upstream of the LTC4368, inrush current into Cin isn't controlled.

The LTC4368 can work down to -40V, so Cin isn't necessary. If you want to filter out brief UV events, add a small ceramic cap to the UV pin.

Ok understood.

So maybe this then with the small gap next to UV pin and a second output capacitor?