And the problem is a capacitive load. There may be some differential ("normal mode") inductance in an EMI filter, but it is likely quite small and "behind" a small capacitor.

 

Just to clarify, solid state devices aren't necessarily switching at zero crossing. It varies from one device to another.

MOSFET based devices don't intrinsically have any zero crossing traits, although they can be driven to switch at zero crossings if so desired.

TRIAC and SCR based devices intrinsically only turn off at zero CURRENT crossings (not necessarily zero voltage.) Switching off at zero current can greatly reduce inductive kick and arcing.

None of the switching technologies intrinsically turn on at zero crossings, but many solid state switches include zero crossing detector circuits which will only switch on at zero voltage crossings. This has many potential benefits, but as mentioned above, it can also have downsides when switching highly inductive loads.

I defer to ebp, crutschow, and other experts here as to whether or not zero crossing will solve your problems. I only wrote this post to clarify the fact that solid state does not necessarily mean zero crossing. If you want zero crossing, you need to specifically make sure that's what you're getting.

(Although I'll admit it's a common enough configuration that you could very easily luck out and get what you need by just trying a few sold state devices at random without examining their internals.)

 

All good info. I looked up the specs for an SSR I already had sitting around, and it claims zero voltage switching.

https://www.amazon.com/Opto-22-Z240D10-Model-Control/dp/B0058V11IG

But what about a NTC thermistor for my problem?