We’ve had this problem several times as well. The first time, we were told that our IR sensor cone was not installed correctly and was therefore interfering with the sensor.
The second time, we were told that the contacts to the build chamber might be loose or dirty and that we should check them, which I personally find the most plausible explanation. The fine powder could interfere with the contacts, so I now vacuum them out regularly.
The third time, we were told to replace the heating rods and the air heater. We did not do this; instead, we stopped using the build chamber afterwards and haven’t had any issues since. We strongly suspect that the problem is actually related to the build chamber itself or its contacts.
@Seunghoon Perimeter melting is usually directly associated with the build chamber. This could be the plug not being fully connected, or broken or dirty pins.
Please take a look at the Melting SLS support page and follow the steps under “Perimeter Melting” to ensure that the plug and the port are clean and undamaged. If the issue persists, I recommend opening a support case so that we can look into the logs and do a deep dive into the printer.
I had this fail too had to replace the socket plug on the fuse. I had multiple Build chambers and had 2 perimeter fails on 2 separate chambers so I knew the issue must be the socket in the fuse. Replaced it and haven’t had that issue since. Absolute ball ache getting the melted plastic off. The build chambers have never been the same since (ie Poorer prints)