Damage to LNK306P
We have a product that uses a PI LNK306 for its' supply. The power (120VAC) is connected to our device through terminal blocks. An installer attempted to connect the power to the terminal blocks live (wire live 120VAC to the terminal blocks) and the LNK306 blew up. I have attached a schematic with a description of the damage. I'm trying to figure out how the part died when almost nothing around it got damaged. Is it possible for the drain to turn on and the device not oscillate? This would cause a short across the unregulated supply and I would expect this kind of damage. Can you help.
Comments
I have attached a .pdf file with a bit more of the schematic and the artwork as well as a picture of the damaged part. The picture isn't great.
Hi,
Some Layout comments:
1.Drain trace seems too close to the Source pin.
2.Drain trace close to BP pin.
One possible reason of the damage could be that there was a conductive residue between the source and drain pins causing a arc when AC input was applied. Arcing could easily pull large amounts of current and heat to damage the insulation of the package and internals.
What is the wire gauge of L1 and Primary winding of T1? Do you see any erosion of the drain and source pins?
I would definitely recommended adding a fuse at the input.
rgds..
The design rules in the artwork call out for 48mil spacing between the drain and source signals. That should be good for over 300V. I don't think there was arcing. Usually, if there is arcing, there are burn marks on the board. In this case, the only visible damage was the top blown off of the LNK306. It looks like the awg of wire for L1 is probably around 40. The current rating is only 300mA. The wire was burned opened where it connects to a pin. It was hard to tell if the pins of the PI part had damage since they were covered in solder when the part was removed. There were no obvious burn marks on the solder mask around the part and the pads were intact and clean. I'm trying to come up with an explanation of what could damage your part. What if the drain got a transient that went negative relative to the source by a bit? The only thing that bothers me about that scenerio is that the aluminum filter caps seemd to be OK.
Hi,
The body diode of the MOSFET turns on when the D pin goes negative to S pin.
Did you see any other burn marks on the drain and source trace when you replaced the PI part? Did you clean the board before replacing the PI part? I am just curious to find out why the failure mode was not repeated when you replaced the PI part.
rgds
The PCB looked clean except the flux from my hand soldering. Remember that this product was suposedly connected to AC while the AC was hot. Two hot AC lines connected to our terminal blocks. I would suspect that the drain going low is less likely than the source going up relative to the drain. I guess it's the same effect. That would explain how the PI part could fail and the caps didn't fail. If the body diode shorted, it would limit the reverse voltage to a diode drop which probablty wouldn't hurt the caps. Then the inductor wire popped and protected the rest of the circuitry.
Seems like a possible scenario. Although it would be difficult to exactly finger point the cause.
rgds
It also looks like both parts that blew up had damage at the same spot, the PI part drain.
Hi,
Never experienced such a situation. Can you please send me the complete schematic showing the bias circuit for the LNK306P and also the Layout snapshot around the IC?
rgds