TOP247Y overtemperaturature shutdown problem in some parts of the Batch "J508 TOP247Y 32952A"

Good morning.

There´are some threads similar of mine, as below:

"Power supply shuts off from time to time" and "Linkswitch-XT LNK364 temperature problem with 2 of 100 chips from one production lot".

I have gotten a over-temperaturature-shutdown about 40 to 50degrees in some parts of the Batch "J508 TOP247Y 32952A". It is very interesting that if I change the part by a new one, the circuit works correctly.

I tested 20 power supplies in an oven at 70degrees. I did the following procedure:

**Heated the entire power supply in the oven at 70degrees for 5 to 10 minutes.
**Then I catch out it from the oven and fed the power supply with 120Vac(it was developed to work from 80 to 240Vac), but it simply do not work. The Top247Y simply do not try to start(I checked with the osciloscope). There is not a start pulse. Nothing.
**When the temperature decreases, the power supply startup and works fine.

Than I bought 10 new TOP247Y from the same batch and supplier, and replaced it in the power supplies that was not working. And them worked well, but 3 of them showed the same problem.

I think it is a problem in this lot. Why some of the parts have the overtemperature-shutdown problem and others not? A large quantity, about 30% has this problem.

I checked TOP247 datasheet. There was a mension about the maximum soldering temperature, time and distance from the terminals to the case. But I respected this concern when I welded that last 10 parts, but the problem repeated in some parts(3 defective parts of 10).

I think that whether the problem was the project(layout, values, etc...) the problem must not went away when I changed the parts.

Awaiting for your answer. thank you a lot.
Eduardo.