Other Parts Discussed in Thread: UCC24612
Hello Ulrich:
Last week there was no internet in this area. So we could not update anything.
Here are the findings on root cause of so-called controller pin shorts.
- #1 Most likely cause of REF pin is that in the PCB, VDD pad & REF pad are 20 mils apart. It is highly possible if not certain that while probing around I inadvertently touched both together.
#2 NTC goes to 2N7002k through a short interconnect. There are 2 power nodes -one on each side of NTC at the PCB edge. Again, cannot rule out straddling the scope probe while NTC was trying to go up. Low probability but most likely happened. One side is 12V for GaN bias set up in the secondary ( pre-biasing UCC24612) & the other side is 12V coming in from an external source ( special case for our set up).
So NEW Rules state that: one can no longer probe around randomly. It should be thought out before power is applied, solder a stub at nodes of interest. Hook probe(s) before PUP.
Another update: Our assembly without controller installed( without BAS126 as well) is thoroughly tested by applying 2usec low rep rate pulse train. I have gone up to 75V & checked all signals such as VSWS, VS, Secondary output, pick up at various nodes around the controller.
We have measured current through the lower GaN using current probe and it looks as expected. But no luck with probes because the gnd lead picks up an awful lot of ringing. So we take the current probe result as true result. The value of the ramp is as expected for Vbulk & 2 usec & Lpri.
It turns out that FB pad was picking up very large signal from VSWS swithcing .. So we are putting Cy caps to prevent such parasitic pick ups
I am with a high expectation that we will get it working in the coming days - but with proto4 assembly.
-robin
