I'm using an LM5085 as an adjustable regulator. My input voltage range is 6 to 14V and my output voltage range is 1 to 14V.
According to the data sheet, Rt should be selected to ensure t_on is at least 150ns at the maximum input voltage. In my case, this works out to be around 7kOhm.
Just down from there, the data sheet goes through a calculation for Rt based on V_in, V_out, t_d, and F_s. Using 300kHz as a switching frequency, 57ns for t_d and my limits for V_in and V_out (14V and 1V respectively), I get a recommended Rt of 9.85 kOhm. However, if I use a different output voltage in the same calculation, I can get values of up to 200k as V_out approaches V_in.
So is it sufficient to use a single resistor value to represent the 'worst case' scenario (e.g 7 to 10kOhm), or do I need to use a digipot and vary Rt as I adjust the output voltage (this wouldn't be difficult...I'm already using digipots in the voltage and current sense feedback dividers to control those two parameters)?
The datasheet also claims that it requires a certain amount of ripple current at the feedback pin to ensure proper frequency performance, and provides a few example circuits to accomplish this. However, in the situation where V_out approaches V_in, this ripple current approaches 0 (it is 0 when V_in = V_out and the FET is held "on"). So...how do I go about generating this ripple current if I want to be able to run the regulator at 100% duty cycle?
On another note...will I run into any stability issues using digipots to dynamically adjust the output voltage and current limit?