I'm fixing a heated water dish, used so that the dogs don't have to eat snow. The previous dish was heated with a wire embedded in what appears to be ceramic, on a foil back on the bottom side of the dish. This failed. Previous experience with this company's dish (Farm Innovators) indicates that 1-2 years is a typical life span for this type of dish.
The rated power for this assembly was 60W. It had a disk thermostat with about a 15 degree F hysterisis in series with the heating element.
It occurs to me that I could replace this heating foil with a handfull of 10 ohm, 10 W power resistors, connected in series, and glued to the bottom of the plastic bowl using thermal glue.
However, this puts the heat in a much smaller space. It's not clear to me if this poses a fire hazard. I'd like to find a guide that gives typical operating temperatures for a given geometry of resistor and surrounds.
In normal operation, the thermostat should be sufficient to protect the bowl and floor, however, I figure I need to design so that failure of the thermostat to open doesn't result in too high a temperature.
I would like to include a temperature sensitive fuse in the circuit. Some that melts at, say 80 C -- well above the operating temperature, but low enough to not catch fire.
Alternate method: Use 32 guage nichrome wire (10 ohms/foot) I would need 20 feet of this with 130v (rural power) to get approximately the same heating effect. Invert bowl. Glue roofing nails to bottom of basin. Pour 1/8" plaster of paris or tile set mortar. Allow to set. Trim nails to 1/8 above mortar. Wind wire from nail to nail in zigzag. Place thermostat. Add 2nd layer of bedding substance.
Nichrome reaction with highly basic mortar?
32 guage wire would puts a lot of heat flux through small area. Would the wire embedded like this overheat?