Swelling RTV-11 potting material in transformers could cause a ferrite to crack. Here's a remedy.
I was involved with high-voltage power supplies in which transformers were wound using 3019-sized ferrite cores. Transformer structures were essentially the following:
Figure 1 Pot core based transformer
Voltages on the order of 1.5 kV were involved so silicone rubber, type RTV-11, was used as a potting material. Assembly was as shown below.
Figure 2 Mounting and encapsulation
We started experiencing cracks of the ferrite cores, so I contacted the core company's tech rep for some technical advice. The fellow came over, we showed him some cracked cores and he listened to our description of what was going on. Unfortunately, his only response was, "Yeah! That happens sometimes."
If there is one thing I really dislike, it's being told that some unacceptable product failure is somehow ‘par for the course.’ This guy was either pretending to be stupid, so he could simply slough off the issue thereby leaving us high and dry while not having to extend any effort to help us in any way, or his apparent stupidity might have been genuine. Of the two explanations, I tend to favor the former.
Since I figured I might need this guy for something else one day, I held my tongue but verbiage that went through my mind was not expressible with any vocabulary suitable for use around small children nor in the presence of polite company.
We eventually worked out for ourselves what was going on.
Figure 3 Core cracking
The RTV-11 potting material was exhibiting a thermal coefficient of expansion. When the material would swell, it would push against the ferrite that was being restrained by the mounting hardware and sometimes, the ferrite would crack. The remedy was to trim the RTV-11 back so that it would not impinge on the ferrite.
Figure 4 Transformer with trimmed RTV-11
By allowing the RTV-11 some room to expand without impinging on the ferrite, core cracking was prevented. Also, while this was indeed a high voltage application, we were fortunate that the approximately 1.5 kV maximum winding voltage wasn't high enough to make corona an issue in the air spaces.
John Dunn is an electronics consultant, and a graduate of The Polytechnic Institute of Brooklyn (BSEE) and of New York University (MSEE).