I recently had an aircraft with a right side window heat showing inoperative with switch selected on.
The EICAS had a warning also.
A test on the Window Heat Control showed only a "Control Unit" fault light. The problem did not follow a swap. The suspect unit tested fine when placed in the opposite position.
There are basically two separate window heat systems. One controller handles the Left Front (#1), the Right Sliding (#2), and the Right Fixed (#3) windows. The other controller does R1, L2, and L3.
Being that there are only two fault lights for each side, FWD for the #1 only and SIDE for #2 and #3, this issue could be caused by either the right side's sliding or fixed windows.
Checking the window assemblies is quite easily accomplished from the controller rack plugs. Each window has a temperature sensitive sensor and one heater coil. Sensor resistance climbs with higher window heat. The 767's sensors came in around 300 ohms with cold windows. Both right side window sensors read good. The only other check to do is to verify the window heater coil to ground. As with most heating coils this reading should be pretty low . The right sliding window came in at 20 ohms..... the right fixed..... 1.5K ohms. Even though the controller did not pick this up, the window is bad.
The #3 window's heater coil ground was verified along with continuity on the "controller to window" power wire. This "does" need to be done as either of them being open would render a window change as pointless.
The controllers handle heat for the front windows slightly different than the sides. For the front window's protection from heat shock, voltage is slowly "ramped up" when power is first applied. The front window receives two phase A/C power. The controller adjusts "B" phase as the means of limiting current. The side windows receive 115 VAC in pulses.
The controller monitors current via a set of Current Transformers on the power output wires. With this aircraft, heat was selected on, but the #3's high heater coil resistance prevented current flow. The controller sensed this, threw up a fault light, and cut the power output. I didn't verify if the sliding window was still receiving voltage or if the system renders both windows inop.
My knowledge of electrical theory is weak to nonexistent. I've always thought of current flow as "power to ground", but with the 767 I've found two instances of using two phases of A/C working against each other.