Hello Trenton,

The problem you are running into is a minimum on-time limit which is caused by the large voltage difference between your input and output voltage. 

The minimum on-time limit is set at 70ns. The closest I could simulate your design is 0.85Vout @ 0.8A  which gave me a duty cycle of 7.01% which at 1MHz is 70.1ns. The datasheet lists that the minimum on-time is max of 70ns and typ at 35ns. Webench uses the maximum rating to ensure the design will always work.

Thanks,

Joseph

Should I not recommend this part for 13Vin to 0.85Vout at 0.3A? 

Since the design with these input/output specs puts the device at this edge case I would probably not recommend it to be used under these conditions as it can't guarantee its working. What application is this needed for as there may be a better alternative? 

Thanks,

Joseph

There is some dependency in a detailed duty cycle equation on the output current; using equation 6 from slyt293 app note Understanding output voltage limitations of DC/DC buck converters. Using the worst case RDson of the FETs (95mΩ HS and 39mΩ LS), Vin 13V, Vout 0.85V, and Inductor DCR of 25mΩ I plugged in 0.3A load current and 1A current to see the impact on duty cycle. At 0.3A the duty cycle is 6.6948% and at 1A the duty cycle is 7.0612%.

Then I used equation 7 from that same app note and plugged in for the same conditions. At 0.3A load the Vout_min is 0.889624V and at 1A load the Vout_min is 0.84208V. They do have a ferrite bead on the output to get lower noise BLM18SG331TN1D which has a DCR typ. of 70mΩ. Therefore the voltage drop would be 0.3A * 0.07 = 21mV. Even with this additional voltage drop we'll still be at 0.8686V at the point of load. This doesn't take into effect a Vin a little bit above 13V, tolerance of output accuracy, or that the load might not be consuming the full 300mA. 

I need another device given the above. I was thinking about TPS62901/2/3 but looks like we'd be in the same boat even though the datasheet only specs the typical Ton_min. We could maybe put it into auto PSM mode and then the Fsw would scale down to ~950kHz. Do you see any issue with using TPS62901? Even though the ton_min max isn't specified in the datasheet, the fact that it has auto PSM mode would adjust for a potential higher ton_min and would scale the Fsw lower in order to still regulate to 0.85V.