190906-1045
S'mise:
The reason to measure the voltage across the load, presumably one low wattage LED, with a high impedance meter, 10 megohm or higher, is to know the approximate voltage across the bulb to cause the glow. This is background knowledge.
Romex cable capacitance is possibly around 20 pfd per foot. That is between the two current carrying conductors in free space ignoring the EGC. With a 100 ft cable that would be 0.002 ufd, or about 1.3 megohms at 60 Hz. 1.3 megs at 120 V = 0.1 milliampere. But it is unlikely that a 120 V energized hot wire is real close to the switch Romex. Thus, less capacitance from some external energized hot wire. Further the Romex cable from the switch to the LED is likely to have two grounded conductors (neutral and EGC) connected back to the main panel neutral and ground bus bars, and we have the switched hot wire, with no power applied, having its cable capacitance to both the grounded neutral and the EGC as a shunt to ground which parallels the LED. Thus, the shunt impedance across the LED and including the LED is relatively low compared to likely stray capacitance from somewhere else. A voltage measurement will provide some idea of the relative relationship.
We need some idea of the leakage current magnitude.
My suggestion of a 15 W bulb as a current shunt to measure leakage current was to provide a convenient resistor to use for the measurement. Unfortunately it changes resistance with applied voltage. Thus, the indirect means to determine leakage current.
At room temperature and no applied voltage a 15 W bulb would be around 70 ohms, prediction from 75 W below. Measured with my Fluke 27 directly across the 15 W bulb I read a stablized 85 ohms, 0.04 V for the 15 W bulb. This reading might be slightly high from self heating, or it could be bulb variance. A 25 W read 46.6 ohms at 0.02 V, and a 75 W read 13.5 at 0.006 V. My recollection is that a 100 W read about 10 ohms. The 25, 75, 100 values correlate. I would have to look at other 15 W units or run a lower voltage test on the existing blulb to determine if a heating effect is present. My previous post means of using the 15 W bulb to measure leakage current cancels the heating effect by doing the current measurement at the same voltage as when in series with the LED bulb.
On a random 9 W 120 V LED bulb it required 0.13 mfd series capacitance from 120 V to cause glow. 0.12 mfd no glow. 0.13 mfd is about 20 k ohms at 60 Hz. 20 k at 120 V is about 6 mA. This is a large amount of capacitance from cables in a home.
A 100 ohm carbon composition resistor would be better for a current measuring shunt. 10 mA at 100 ohms is 1 V. The reason to use an external shunt and a voltage measurement instead of directly measuring current is to protect your meter.
.