3. Testing
This post continues from Part I by showing how the power supply was tested.
The only issues encountered during assembly was an incorrect cap on the external loop. C1 from 1uF to 100nF.
First a preliminary simulation using TI web-bench was done .
As a rule of thumb all power supplies are simulated using either LTSPICE , Symmetrix/Simplis or an other SPICE based application tailored for SMPS.
The main idea is to check if the power supply is stable. This in practice is done by breaking the loop and injecting a small 10mmV sine wave and checking the BODE plot (gain and phase margin)
In this case however there is no need for loop compensation since we are using a regulator with internal loop compensation with an integrated MOSFET so the only components that need to be specified are the output voltage resistor divider and the LC output filter. While these are important for the output voltage they are much easier to select compared to a full loop compensation which would require special equipment such as a Venable machine, signal generator, 2-3 power supplies, and a much better scope.
The following video shows testing the 5V power supply.
a) Turn on voltage test
b) shut down test
c) Power supply voltage turn on transient
d) Current ripple . This test requires a current probe so it was skipped.
e) Overvoltage shutdown 50V-60V.
f) Over current shutdown
g) Voltage ripple (23mv)
The same board also has a 24V power supply.
Normally one would implement the power distribution network by daisy-chaining the 5V and 24 V from 30-60V. in this case both are sourced from the same input.
Figure1. 5V power supply output.
As a last note power supply projects emphasize the need for : a) simulation and analysis and b) a well equipped tool-bench.