Most inexpensive X10 devices have transformerless power supplies, which use a capacitor to drop the line voltage to the low level needed by their electronic circuitry. The charging current is reactive, which is not in-phase with the applied voltage. The Kill-a-Watt has two ways to measure power consumption - watts and VA.
Watts is the "real" power that you pay the electric company for. VA (volt amperes) is the average of the real-time multiplication of voltage and current over the entire AC cycle. Purely reactive loads will draw current charging up at one point in the AC cycle, and dump that energy back to the poweline elsewhere in the cycle, resulting in no "net" power consumption. However significant current can be drawn during the charge and discharge.
The amount that VA differs from watts is a function of the power factor of the device. When the power factor is 1.0, the VA and watts will be essentially the same. At lower power factors, the numbers can differ by large amount. For a Maxi Controller, the Kill-a-Watt reads 1.0 and 10 for watts and VA respectively. Even though the Maxi Controller label says it only consumes 2.5W, the low pass filter in the XTB must deliver the same current as if a 10W load was plugged in.
You can think about this in another way. If you stick a big capacitor - say
2.2uF - into an AC socket, it will pull about the same current as a 1200 ohm resistor. If that were "real" power (watts), the capacitor would dissipate about 12 watts. However, the capacitor won't even warm up because the current is "imaginary" (90 degrees out of phase with the applied voltage). I just verified this with my own Kill-a-Watt. It measured 0 for watts, and 11 for VA with a 2.2uF capacitor.The power supply in a X10 transmitter functions a lot like that capacitor. The only "real power" that the Kill-a-Watt measures is that actually consumed by the module electronics.
It the case cited earlier wherein the XTB low pass filter was overloading by removing one of the 5 loads, that one load had a power transformer. So, its "imaginary" inductive current was opposite that of the other X10 transmitters, partially canceling out their effect. When that load was removed, the current pulled by the 4 remaining capacitive loads (about 40VA) exceeded the rating on the XTB low-pass filter inductors. This is similar to the issue that the power company deals with by placing capacitors in their distribution network to balance out the inductive reactance from the various motor loads.
I realize that this may be a difficult concept for some. Hopefully, the capacitor example above will help you understand the issue.
Jeff