This post is related to the problem that ZetaFunction is having with his OBC failing to start up the LV supplies on the bottom board.
i have been puzzled by how the feedback worked to regulate the ~16V created by the big secondary on the transformer T302, so i have been doing some research and made a simulation based upon notes from the datasheet. It's hard to believe all the stuff they crammed inside that little 5-pin voltage regulator device.
The Red LED represents the photodiode of the photocoupler.
The sim is switching the FET at 75kHz with a 10V pulse at about 10% duty cycle. i could vary the duty cycle to raise or lower the buss voltage; i did it just enough to understand how it works and then stopped--didn't try to tune it for a perfect match as i had to manually change duty cycle.
On the OBC bottom board when the photodiode comes on, it pulls down pin 2 of the NJM2369 chip, which cuts off the PWM to the FET. This occurs when the internal NPN transistor is turned ON by the internal opAmp when the rail voltage goes above the desired set point and raises the voltage across the 3.9k resistor (call it R2) which is the non-inverting junction. Then when the FET is OFF, the buss voltage begins to bleed off until such time that the photodiode is no longer ON and as the R2 voltage drops below the 2.5Vref at the inverting input. When the photodiode is OFF, then the voltage at pin 2 of the NJM can rise back up to cause the PWM switching to occur again. So this device provides a feedback loop to watch and control the 16V buss.
i have been tracing the other photocouplers looking for some connection that might be causing your issue, but so far no luck.
i have looked upstream from the 16V regulator at the "loads" but haven't figured it out other than if something in the waffle plate could be pulling it down. The "ground" of the isolated 16V supply is floating with respect to the 12V battery ground. It actually connects the negative side of the C845 to the negative side of the HV DC from the AC rectifier and the negative of the 3 big electrolytic caps. The positive side connects to the rail of the push-pull drivers for the PFC switching FETS in the waffle plate. i'm thinking that maybe something else is causing the LV supplies from starting, but i'm grasping at straws...
[edit: the screenshot clipped some of the numbers, the cap voltage is 19.07 and the resistor at the top is 1Ω ]