Other Parts Discussed in Thread: UCD3138
This is a question that ties UCD3138 controller for our system. The power will be rectified by 2 LMG3410 as shown in the earlier msg.
Can you review this attached pdf & see if this makes sense?
Here is the thinking process that has gone on here.
Since GaN does not have "body diode" that begins the startup build of output voltage even in absence of sync rectifier drive signal, we have to adopt a little aggressive method yet without any operational hazard.
In our power stage, we know by computation/modeling approx sync turn on - when it starts & approx duration. ( typ 1 usec)
Then , why not force this pulse on the sync switches in order as shown in the attached waveforms? And when the controller decides steady state, it fine tunes the drive signals.
For this to happen, we must sense DC current at appropriate places which we can do using AA024 ultra miniature 1 Meg HZ sensor from NVE. Interestingly, in this manner of sensing will make the sensor delay a moot issue.
Because of this sensing, we derive the sync ON pulses for steady state which will be marginally different from forced conditions at startupSYNC_RECTIFIER_TIMING_DERIVATION_OUTPUT_SENSING.pdf.
What would you think of this strategy?
Would we be able to provide the sensed signals to the UCD3138 & get the pulses from one of the many PWM outputs? if these are too narrow, we can feed a comparator output as hi/lo signals as well. There is no absolute need for UCD to deal with the analog fast sensor signals.
